Electric utility forecasting and planning solution for grid modernization
For electric utilities who need to support grid modernization initiatives with load forecasts that meet top-down regulatory requirements and bottom-up DER impacts, and who want a transparent, proven, modular approach built on a scalable analytics platform, Corios has developed our Lightning solution.
Corios Lightning is a distribution planning and forecasting solution that provides a 10-year hourly forecast of MW demand for every substation and feeder on your grid, adjusted for economic growth, load growth projects, capacity transfers and DER adoption.
What makes Corios Lightning unique?
Other solutions emphasize glitzy “AI” features, but we focus on solving the hard problems: giving you a solid foundation to make mission critical decisions affecting your grid and your customers.
- Hard-fought lessons learned about data quality, stability, reliability and believability have been baked into the Corios Lightning solution. This helps you avoid paying the penalty for learning these lessons all over again.
- Corios Lightning is a production-ready software application, including a robust set of functional processes, user interface, data model and dashboard reports.
- Lightning is scalable and reliable. We have field tested and received acceptance of Corios Lightning in production at Southern California Edison, one of the country’s most forward-thinking and progressive utilities facing some of the most advanced set of challenges posed in the California market.
- Lightning addresses all the functional expectations for California regulatory requirements, which other states are expected to eventually follow.
- Corios Lightning is transparent. Unlike other solutions that are compiled, black-box and one-size-fits-all, Corios Lightning is an open, white-box and customizable solution.
- Lightning is modular. If you want to revise or customize part of the Lightning modules, we support that and will tailor to your specifications.
more about our Lightning solution
Client Case Study: Southern California Edison grid optimization
Learn how Southern California Edison (SCE) is successfully planning for challenging California carbon legislation impacts and modernizing its electric grid.
Read the Case Study →
Corios Lightning solution video
Watch a six-minute video that walks through the day-in-the-life of the distribution planning engineer who uses Corios Lightning to plan grid modernization projects.
Watch the video →
Corios Lightning overview RedPaper
Read a detailed description of the Corios Lightning solution.
Read the Corios RedPaper →
Corios Lightning case studies
Read more about the challenges, solutions and results of Corios Lightning clients in the electric utility industry. To see our complete database of case studies, click here.
|ID||Industry||Client business type||Analytics challenge||Business problem||Challenge||Solution||Results||Corios Solution|
|1||Banking||Auto lending||Revenue forecasting||P&L forecasting and simulation||The manager of an automobile lending portfolio for non-captive dealers nationwide engaged Corios to optimize their profit & loss (P&L) pricing model. Their business-as-usual model utilized only nine risk tiers, and required roughly 15 minutes to calculate the net present value margin contribution and return on assets (ROA) for each scenario of a single set of assumptions. These limitations kept the pricing team from developing more sophisticated pricing strategies and optimizing for the ideal price structure on a dramatically more segmented portfolio of dealers and loan prospects.||Corios developed an automated approach for pricing P&L modeling that integrated the client’s expanded hierarchical segmentation strategy and underlying warehouse records. The new model allowed their pricing VPs to create thousands of segment-specific P&Ls in under a minute. For each P&L, we developed an optimization approach to identify the ROA-maximizing discount rate on each segment of loans, effectively running hundreds of scenarios on each segment. A simple user interface was implemented, allowing the non-technical financial analysts in the team to run their own assumptions and scenarios without technical assistance.||The financial analysts for the lending portfolio are now able to optimize pricing terms at a more precise, highly segmented point of entry in a dramatically reduced time frame (from hours to seconds). This P&L model enhancement is conservatively estimated to have improved the client’s ROA by several percentage points at the portfolio level.||Forte|
|2||Credit union||Auto lending||Customer profitability||Collections scorecard and strategy assignment||The auto loan servicing division of this financial services company had a "plain vanilla" collections strategy where every debtor account was handled using the same scheduled approach, regardless of the asset value, the credit risk, the asset recovery, or the availability and effectiveness of recovery resources.||Corios built a mathematically optimized strategy to maximize the risk-weighted returns on collections activities by assigning the ideal delinquency & recovery strategy to each of roughly 170,000 accounts in the loan portfolio.||Incremental benefits for the solution estimated at $2 million in incremental net receivables per month.||Harmony|
|3||Banking||Banking services provider||Revenue growth||ATM transaction forecasting and optimization||The ATM transaction forecasting team of this banking services provider could not accurately or precisely project the daily cash-in and cash-out volumes for each of thousands of ATMs for their clients. This prevented them from being able to provide strong service levels for ATM servicing crews and vehicles.||Corios built a series of econometric time series models to forecast the expected interval of daily cash-out transactions per ATM location for a financial services systems provider. Models accounted for recent historical volume, cyclical and payroll cycles, seasonal and holiday events and geographic proximity to other ATM locations.||The forecasts Corios developed for the client met and surpassed client expectations for the forecast quality, especially compared to business as usual.||Tempo|
|4||Credit issuer||Banking services provider||Analytics modernization||Real-time model deployment & model lifecycle reporting||The transaction servicing division of this banking services provider was falling behind its competition because it was unable to deliver compelling real-time offers at the point of sale for tens of thousands of their POS merchants, in part because they lacked the capability to score transactions in real time against five to ten analytic models built outside of the POS platform.||Corios replicated three predictive scorecards originally developed by a banking services provider, and then deployed the scoring routines for these scorecards in a format that provides real-time scoring via electronic point-of-sale capability. We augmented the existing models by delivering reject code processing and score binning for decision-making. Finally we developed a series of model specification reports that automated the existing manual process of spreadsheet-based model documentation.||The deployment time frame for making these models in production was less than a month, and the run time per model was thousands of customers per second. This met and surpassed the client's expectation for delivering this capability to their own customers.||Rosetta|
|5||Investments||Brokerage||Cost reduction||Call center demand forecasting||The designated benefits division of this business financial services provider could not project its call center staffing accurately enough to maintain service levels for call handling, due to high turnover.||Corios developed a series of econometric time series models to forecast the expected monthly inbound call volume for the business benefits division of a leading investment brokerage, by client, contract type and relationship history length. Models took into account contract renewal events, publicly-traded client equity share prices, macroeconomic indicators and seasonal events.||Compared to business as usual, our models reduced overall portfolio forecast error from 15% to 6% for the 3-month planning horizon, and reduced operational risks via the estimated annual cost set-aside for projected call center agent hiring and training costs by over $2 million.||Tempo|
|6||Investments||Brokerage||Revenue forecasting||Equity price forecasting||The economic forecasters for this retail brokerage could not accurately develop fine-grained forecasts of portfolio performance for the securities and portfolios in which their clients invested.||Corios developed an approach for our client to utilize powerful econometric and time series techniques that enabled them to build better accuracy and precision, as well as deep granularity, into their perrformance forecast scenarios.||The client's forecasting team increased their coverage of client portfolios by a factor of 10 or greater, and reduced the time required to generate those forecasts to less than a single day.||Forte|
|7||Investments||Brokerage||Customer profitability||Financial advisor contact targeting||The relationship management team of this money management firm could not attribute centralized marketing and distributed financial advisor relationship manager activities to changes in assets under management. The client sought to better target the most effective relationship treatments to grow AUM.||Corios developed a behavioral segmentation strategy and a pair of predictive scorecards. The segmentation was comprised of advisor attributes, past sales and redemption patterns and transaction event sequences; the predictive scorecards focused on increasing the targeted response rate by scoring on transaction volumes and segmentation-based factors.||The increased response rate from the scorecard exceeded business-as-usual predictive accuracy by over 20%.||Harmony|
|8||Investments||Brokerage||Analytics modernization||High-volume model scoring||The marketing analytics department of this retail brokerage wanted to significantly speed up the processing time of their predictive scoring algorithms for customers, so that their financial advisors could compete on the same cadence with other leading investment firms who could update their customer strategies on a daily basis.||Corios developed a pair of predictive scorecards that estimated likely adoption of a bond product offering, and the expected assets under management volume for a group of several million accounts. We then deployed those models into the brokerage’s enterprise warehouse as a pair of high-performance scoring algorithms.||Corios' predictive scoring approach reduced the time to scoring each model on 5 million accounts from a few hours to as little as a few seconds. This allows our client to refresh their view of the customer and the most effective marketing strategies to be refreshed nightly instead of weekly, allowing them time to more effectively compete for their customers’ investment interests.||Rosetta|
|9||Investments||Brokerage||Customer profitability||Sales leads assignment optimization||This retail brokerage wanted to refresh their client management and modernize their asset growth strategies by assigning the optimal lead for thousands of financial advisors and call center agent across 2+ million accounts.||Corios developed a mathematical optimization routine for our client that maximized weekly sales NPV by assigning the ideal sales lead per account and financial advisor across 2+ million accounts and thousands of advisors, refreshed on a daily basis. We incorporated a multiple-week planning horizon, and multiple contact channels. Business rules used as policy constraints included advisor licensing tied to sales offer, advisor calling capacity, cost per contact and preservation of lead volume over days in the planning horizon.||The incremental benefits compared to the prior approach for lead assignment was in the millions of dollars per multiple-week planning horizon.||Harmony|
|10||Banking||Corporate bank||Compliance and Governance||Credit and market risk model validation and lifecycle management||A US corporate bank client needed to satisfy regulatory requirements for their predictive model catalog pursuant to Basel II.||Corios developed a model validation and lifecycle management platform for our client. Specifically, we automated the model inventory process, incorporating their model review workflow and delivered scheduled notifications to staff for tasks on the critical path.||Our client was able to pass compliance audits with the OCC and Federal Reserve in part through satisfying the requirement for their predictive model catalog enterprise-wide.||Metronome|
|11||Credit issuer||Credit card issuer||Revenue growth||Balance attrition prediction||This credit card issuer was concerned that customers would reduce their revolving credit balances in favor of shifting these balances to other credit issuers. They wanted to avoid the loss in net interest income tied to this threat.||Corios developed a class of predictive scorecards to target accounts at risk of a revolving balance reduction through a gradual or immediate pay-down event. We defined the complex event that characterizes types of pay-down, and built 4 scorecards to predict leading indicators of balance retention by risk type.||The issuer doubled the proportion of saved at-risk accounts in the top two deciles of the targeted account population, compared to the de facto strategy.||Tempo|
|12||Credit issuer||Credit card issuer||Customer profitability||Initial credit line assignment optimization||This national credit card issuer wanted to set the ideal credit line for each new account at the point of booking, in order to maximize the profitability of these new accounts.||Corios developed a mathematical optimization routine for our client to maximize risk-weighted net income by assigning the ideal initial credit line to each account at booking. Deliverables included behavioral segmentation tied to credit risk, utilization, balance and limit, and development of predictive scorecards for likely net income and gross credit loss. The optimization routine accounted for policies on total credit exposure, total expected loss, active rate and utilization, in order to calibrate the optimized assignment to current portfolio performance.||The best-performing optimization scenarios projected an incremental net income contribution between 200% and 400%, and a margin-to-revenue ratio increase between 60% and 240%, compared to the business-as-usual performance on the historical accounts.||Harmony|
|13||Credit issuer||Credit card issuer||Analytics modernization||Marketing analytics platform modernization and targeting models||This large credit issuer needed to modernize their entire customer marketing platform.||We implemented a marketing analytics platform for our client, consisting of linkage to the enterprise data warehouse, development of several targeting models, scoring integration into the warehouse, and development of automated reporting on campaign performance. We trained over 30 staff in the ongoing design, maintenance and expansion of the platform.||Through this marketing analytics platform modernization, our client was able to broaden the number of campaigns, predictive scoring models, and marketing optimization scenarios, and ultimately improve their speed to market and their campaign performance for all campaigns.||Rosetta|
|14||Credit issuer||Credit card issuer||Revenue growth||Prescreened acquisition line amount optimization||This national credit card issuer wanted to set the ideal credit line for each new pre-screened applicant at the point of making an offer of credit, in order to maximize the risk-weighted profitability of these new accounts.||Corios developed a mathematical optimization routine for our client that assigned the ideal credit line for each newly acquired account to maximize risk-weighted earnings before taxes. Compared to the business-as-usual approach, Corios' approach increased the profitability of line assignment by expanding the line assignment to the account level while also enforcing more demanding credit policy rules.||Using our strategy, the credit issuer's expected annualized incremental EBIT for a portfolio of 2.8 million accounts was projected to exceed $80 – $120 million.||Tempo|
|15||Credit issuer||Credit card issuer||Revenue growth||Recurring billing offer optimization||This national credit card issuer wanted to maximize the response rate for a new marketing offer that offered targeted incentives to accounts who have recurring monthly payments to the same merchants, including telecom, garbage, insurance, and so on.||Corios developed a series of predictive response scorecards and a mathematical optimization routine for our client. Business rules were designed to control the outbound mailing and inbound incentive budgets and also provide sufficient capacity in all test cells to permit robust response rate performance testing.||On an eligible universe of 8.6 million accounts, we targeted slightly more than 700 thousand offers, generating $700 thousand in income for every month over the customer's subscription in the program, for a one-time direct marketing investment of $695 thousand.||Tempo|
|16||Credit issuer||Credit card issuer||Revenue forecasting||Sales forecasting||This national credit card issuer wanted to develop a more accurate and more granular revenue forecast so that they could attribute their marketing campaigns' effectiveness on revenue contribution.||Corios developed a monthly revenue forecasting system for our client to project their revenue drivers (finance charge, balance transfer, cash advance, late fee and interchange) 12 months out, as a baseline for targeting and measuring marketing campaign impact. Unlike their business as usual forecast, the enhanced forecast portfolio was highly segmented across product, spend behavior, credit risk, purchase APR, wallet share, vintage and merchant spend dimensions, and accounted for comparisons across macroeconomic response, behavioral migration and acquisition strategy scenario planning.||The revenue forecasts yielded by this system helped senior marketing leadership to improve the accuracy of their baseline revenue forecast, and to attribute campaign investments to drive profitable growth.||Forte|
|17||Credit issuer||Credit card issuer||Revenue growth||Sales migration prediction||This national credit card issuer wanted to identify the leading drivers of attrition risk linked to active-spending cardholders who are likely to slow or end their spending on that issuer's credit card, resulting in income loss.||Corios augmented our client's existing predictive scorecard to improve the leading-indicator detection rate of active-spending cardholders who might migrate into inactive behavior. We developed novel predictors tied to spend transaction behavior and transaction sequence models which resulted in significant model performance improvement.||The Corios strategy improved accurate identification of at-risk accounts by over 100% on transactor accounts and by more than 15% on revolver accounts in the top decile of targeted accounts. Our predictors also revealed the customer behavioral drivers to which personalized offers could be used to improve saving the account.||Veloce|
|18||Credit issuer||Credit card issuer||Revenue growth||Credit line increase optimization||In three separate engagements, these national credit card issuers wanted to improve the profitability of their existing cardholder profitability by re-assigning the ideal credit line for each customer.||Corios designed and implemented mathematically optimal credit line assignment strategies. These strategies accounted for modern credit risk policies on total and at-risk balance exposure, percent and absolute increases in line per account and per credit risk tier, and calibration of the optimized assignments to existing strategies.||For these three engagements, the projected incremental net income ranged from over $80 million annualized (on 1.5 million accounts), $5 million annualized (on 350 thousand accounts) and $60 million annualized (on 3.5 million accounts).||Tempo|
|19||Credit issuer||Credit card issuer||Loss forecasting||Loss forecasting||In three separate engagements, these banks needed to dramatically improve the approaches they used for forecasting delinquent and charged-off account behavior, pursuant to regulatory requirements and sound business strategy.||The Corios strategy included account volumes acquired and retained, outstanding balances by current and delinquent status, and gross and net charge-off volumes. Forecasts provided for flow rates across delinquent bins, projection of new and immature vintage curves, sensitivity to projected macroeconomic stresses, and management reporting.||Each bank was able to improve the accuracy of their credit exposure forecasts, and also identify the drivers of potential exposure to which they could change their credit policy and avoid adverse outcomes.||Forte|
|20||Energy||Electric utility||Load forecasting||Peak load forecasting and weather normalization||A prominent summer-peaking electric utility client faced challenges with system-wide daily peak load forecast accuracy in the hottest periods of the year.||Corios developed an econometric time series forecast model that incorporated sophisticated predictive drivers of daily peak load, including peak weather events and cyclical demand trends, that reduced forecast error from over 3 percent to under 1 percent for the highest peak load days.||The Corios strategy allowed the utility to fine-tune their capacity reservations to meet peak load demands, thereby reducing the costs of maintaining contracts for more capacity than required.||Lightning|
|21||Energy||Electric utility||Load forecasting||Demand and load forecasting||For more than 40 electric power, natural gas and water utilities, Corios determined that these clients had substandard capability to forecast and react to changing capital investments for infrastructure, as customer demands for the utility's service changed over time, particularly in response to energy and water conservation measures.||Corios developed a demand forecasting system that combined building construction and renovation forecasts, appliance and equipment unit demands, demand-side conservation measure impacts, and building and equipment lifecycle projections across multiple economic and regulatory scenarios. Forecasts were segmented by climate zones, rate classes, building & equipment types and vintages.||Corios strategy enabled policy makers in improving the forecast accuracy and delivering analytic insight related to customer demand in medium- and long-term horizons to drive regulatory, pricing and demand-side program design.||Lightning|
|22||Energy||Natural gas utility||Customer profitability||Geo-targeting sales and marketing application||This natural gas distribution company needed to overhaul their customer acquisition strategy, which had been merely to answer incoming phone calls from non-customers and facilitate conversion sales cycles.||Corios designed and implemented a geo-targeting database and sales targeting system. We consolidated assessor parcel and building data with the gas LDC’s distribution engineering digital maps and customer database. Geo-location algorithms identified the distance of each home to the nearest gas main, and we gave sales representatives the ability to build their own custom digital maps as they prepared for their daily visits to targeted neighborhoods. Conversion and cross-sell response probability models were used to score houses and neighborhoods for the most profitable targets.||The Corios sales targeting system enabled the client's sales teams to identify residential conversion and cross-sell targets more profitably via enhanced information about each building.||Harmony|
|23||Energy||Electric utility||Load forecasting||Optimized labor allocation model||This electric utility needed to reduce the massive cost structure tied to building and maintaining their electricity transmission and distribution network.||Corios optimized the company's labor allocation to construction and maintenance projects for the transmission and distribution system. We forecasted construction and maintenance demand in terms of pole-miles and crew FTE by state and month, and then developed a mathematical optimization routine to allocate the most cost-effective resources to the demand given availability of in-house and contract crews. The optimization routine accounted for in-house staff hiring, training and retirement projections, overtime costs and productivity, and the variable costs of alternative crews and their equipment.||The Corios strategy provided the client with the ability to shave over $200 million annually from their operating costs.||Lightning|
|24||Credit issuer||Credit platform||Revenue growth||Merchant transaction anomaly detection||This global payments processor needed to determine the incidence and severity of anomalous credit transactions for every individual merchant company and DBA in the US on a daily basis. Their existing business rules created many false positive reports and wasn't used by merchant relationship managers because it wasn't trustworthy.||Corios developed an daily merchant-level event detection routine for our client. This routine flagged merchant-specific processing days that were outside the expected range for that merchant, given seasonal, cyclical and administrative transaction signature.||The Corios approach outperformed the de facto approach by increasing true event detection by 30% and cutting the false positive rate by nearly 50%.||Veloce|
|25||Credit issuer||Credit platform||Revenue growth||Targeted incentive campaign scoring||The cardholder loyalty group of this global payments processor wanted to launch a nationwide loyalty and spend stimulation program, targeting incremental spend at selected merchants where the card had not been used by a cardholder in the past 6 months. However, no platform for predictive targeting, segmentation, campaign management or measurement existed to support this program.||Corios built and ran the campaign targeting, scoring and performance measurement process for our client's targeted merchant mail campaign over an 18-month period, including over 20 merchants in the program. We developed transaction-based behavioral segmentation and scoring routines to maximize incremental credit card spend that routinely outperformed standard targeting approaches such as category-specific spending indices.||In one such campaign, compared to the random-selected cell, our targeted segments delivered a $6.9 million projected incremental spend on a marketing budget of $1 million.||Veloce|
|26||Banking||Retail bank||Customer profitability||Cross sell marketing campaign optimization||The rewards program of a diversified retail bank wanted to enhance the omnichannel execution of their rewards behavior targeting program. The business as usual program's targeting was ineffective and resulted in relatively flat response rates. In particular, they wanted to make better use of arbitration between call center and email offers.||Corios designed a simulation and optimization routine for this client that assigned the ideal product offer customers via a web and call center-based cross-sell engine. Business rules for the optimization routine included offer eligibility, response scoring, order presentment and cycling based on past contact and response history, and operational planning constraints.||Expected incremental dollar-weighted ROI compared to business as usual were conservatively projected to exceed 15% over business as usual (equivalent to roughly $2 average incremental response per customer) on a sample of 10 thousand accounts per day.||Harmony|
|27||Banking||Retail bank||Revenue growth||Deposit cross-sell campaign test design||The direct deposit (checking and savings) cross-sell team of this large bank needed to enhance response rates to checking upsell programs. They wanted to learn how to do test-and-learn campaigns, but were stymied since their response rates were measured in the low basis point range.||Corios trained the client's team on the practices of experimental design and campaign cell sizing applied to a deposit account up-sell campaign. We also showed them how to implement these sophisticated test-and-learn strategies in their campaign management platform.||Through our advisory, client staff became well versed in the trade-offs related to single-campaign versus multiple-campaign tests, particularly important in the context of very low response rates and finite eligible marketing populations.||Tempo|
|28||Banking||Retail bank||Revenue growth||Deposit transaction forecasting||The retail branch planning team for this large bank wanted to improve the precision and accuracy of predicting daily deposit transactions in each branch. They struggled with staffing and branch location planning as the transaction volumes varied over the course of the month, season and year.||Corios built a series of econometric time series models to forecast the expected volume of daily deposit account transactions by branch. Segmentation of the forecast models took into in-branch account segmentation, account branch size, type and geographic location, as well as seasonal and holiday events.||Corios' transaction volume forecasts substantially outperformed the bank's de facto forecasts, enabling them to improve capacity planning, and to shift branch staff to the locations to balance their demand.||Tempo|
|29||Banking||Retail bank||Customer profitability||Cross sell marketing campaign optimization||In two separate engagements, these banking clients needed to optimize their omnichannel, cross-product cross-sell campaigns. They struggled with how to balance the competing demands of channel capacity, product sales goals, and customer contact fatigue, not to mention, helping the bank hit their income growth goals.||Corios developed mathematical optimization routines to help our clients maximize the net income generated by monthly cross-sell campaigns across multiple contact channels and banking products. The optimization routine accounted for expected response rate, gross margin, expected default rate, variable cost and likelihood of contact by channel, and rolling contact policies by account. Business rules defined in the offer optimization strategies maintained either outbound offer volumes or direct and indirect budgets, compared to historical campaigns.||Corios' strategy for cross-sell optimization expanded the incremental ROI (compared to business as usual) from 200% to over 600% depending on the constraints used in the scenarios, resulting in millions of dollars of increased cross-sell income each month.||Harmony|
|30||Banking||Retail bank||Analytics modernization||High-performance scoring||The online banking division of a large American bank struggled with using predictive scoring in their web and mobile banking platform because the de facto platform couldn't adapt to the service levels required of their high transaction volume.||Corios replicated the bank's predictive offer selection scorecards and embedded them directly into the bank's transaction database, thereby substantially reducing the scoring process time.||Corios' in-database scoring strategy reduced score processing time by a factor of 10, from roughly 30 minutes to score 1 million accounts to under 3 minutes.||Rosetta|
|31||Banking||Student lending||Revenue growth||Loan pricing optimization||The risk and pricing team of an unsecured lending institution was struggling with how to offer differentiated pricing that adapted to applicant creditworthiness driven by more than just the FICO credit score. The lending organization had to determine how to meet portfolio-level risk policies while also offering better lending terms for more creditworthy applicants.||Corios developed a mathematical optimization routine to assign the risk-weighted margin-maximizing price per account at origination for our client. The optimized pricing routine accounted for the lender’s policies regarding approval decisions, lending and default exposure, loan term eligibility, price elasticity of acceptance, behavioral scorecard-based default rates, and simulated time-to-default estimates.||Compared to the business as usual strategy, Corios' strategy delivered an incremental 24-month NPV benefit between $5 million and $20 million (representing a 25% to 100% percent improvement) for a pool of roughly 200,000 loan applicants.||Tempo|
|32||Banking||Student lending||Risk mitigation||Student loan payment default||The servicing division of a student lending institution was struggling with accurately predicting the default rates of loans on their books. This was a new initiative for the servicer, who had been reliant on generic credit scores for their servicing strategy.||Corios developed four behavioral scorecards to predict the likely account default for a student loan servicer, one for each maturity level of the loans in the portfolio.||Corios' credit risk strategy delivered substantially stronger predictive accuracy of customer default rates; our scorecard made accurate prediction of future loan defaults on 75% to 85% of customers depending on maturity, and reduced false positive rates to between 15% and 22%.||Tempo|
|33||Banking||Student lending||Risk mitigation||Credit risk scorecard||The risk and pricing team of an unsecured lending institution wanted to develop a next-generation scorecard based on their own customers' attributes and performance, instead of relying on a syndicated model provided to them from a vendor.||Corios implemented a credit risk scorecard development platform for a lender specializing in unsecured lines of credit, and trained the lender’s credit risk analytics team in the use of the platform.||Based on the adoption of our credit risk development and scoring platform, our client subsequently developed a series of acquisition risk scorecards that substantially outperformed credit bureau-based score products, which became a standard for the lender’s credit policies.||Tempo|
|34||Energy||Power wholesaler||Analytics modernization||Code audit of power forecasting and settlement system||The energy load forecasting and scheduling division of a US power wholesaler firm struggled with an aging forecasting platform for which only a single analyst was able to maintain. This created substantial key person risk, leaving the entire firm's book of business at risk.||Corios audited and documented the firm’s load forecasting and scheduling system, including linkages into Oracle MDM (Lodestar) used for operational settlement. We documented the logic flow, data flow, and the models used within a sytem utilizing appoximately 10,000 lines of code. We Identified system management operational risks for senior leadership, and recommended solutions including change management, training, and audit preparation.||The firm adopted all our recommendations, and implemented the analytic platform modernizations that we designed. The key person risk evaporated and our client was able to bring on new team members to tailor and maintain the power load forecasting and scheduling platform.||Rosetta|
|35||Banking||Retail bank||Risk mitigation||Financial crime compliance analytics||A retail and commercial bank client faced growing challenges related to routing of BSA-AML alerts to examiner and SAR queues while scaling their labor pool to handle the workload.||Corios developed a financial crimes detection model suite for our client, which contained two models, one to detect the presence of transactions that are highly likely to be financial crimes (based on similarity to patterns of past known criminal activities), and a complementary model to score transactions very unlikely to be criminal activities. The scoring routines used party- and transaction-level attributes, as well as time-based transaction patterns and sequences, to separate true and false positive matches on month-ahead outcomes.||Corios' BSA-AML criminal activity model captured roughly 40% of true positives in the top two deciles of its scores, and 30% of the false positive records in the second model’s top two deciles. This model performance allows the bank to route alerts tied to transactions to dedicated queues, and to focus the examiner pool on the remaining “middle of the road” alerts that have the greatest requirement for manual review. We migrated the models into test and production in two months’ duration.||Veloce|
|36||Banking||Retail bank||Risk mitigation||Fraud and underwriting analytics||The credit risk team of this short-term lender wanted to differentiate application-level loan terms for their customers, but didn't possess the capability to perform the through-the-door predictive scoring and decisioning in real time.||Corios developed a 2-year-forward analytics roadmap, and implemented several analytic models to detect likely fraud and to provide loan underwriting and pricing terms. We produced model performance validation reports on these models, and implemented their scoring routines for real-time execution through their point-of-sale platform.||The credit decision platform that Corios built enabled our client's front line stores to provide instantaneous line amount and pricing terms for their customers for thousands of customers every day.||Tempo|
|37||Insurance||Insurance carrier||Revenue growth||Marketing acquisitions models for response, underwriting and payments||This life insurance carrier needed to increase the profitability of their customer acquisition strategy, but struggled with how to leverage data-driven analytics to define such a strategy.||Corios developed marketing acquisitions predictive models for our client, including likelihood of response to an acquisition offer, likelihood of underwriting approval for the applicant, and likelihood of adoption of the approved policy by the applicant. Corios then identified the optimal acquisitions targeting strategy.||Corios' customer acquisition strategy for this client more than doubled the profitability of customer acquisitions (measured by net value per paid account divided by cost to acquire) to 66 cents per customer, up from the business-as-usual payout ratio of 28 cents per customer.||Tempo|
|38||Insurance||Insurance carrier||Risk mitigation||Analytic model validation and performance tracking||The underwriting team of this property and casualty insurance company needed assistance in modernizing their analytic model processes in order to meet and exceed global guidelines on model transparency and auditability.||Corios developed business processes for our client to deploy and track model assets, track team workflows on model creation and review, track the ongoing mathematical performance of their models, and to support model transparency and governance goals for the underwriting function. We also trained the team in how to import new models into the tool and how to introduce new users to the process as they join the team.||Our client substantially reduced the time they allocated to model tracking, audit support and documentation, so they could refocus more time on the creation of new and updated model assets.||Tempo|
|39||Banking||Retail bank||Customer profitability||Cross sell marketing campaign optimization||The cross-sell marketing team of this bank needed to modernize their marketing campaigns so that they could grow income while also balancing customer contact strategy, channel capacity and bank product sales goals.||Corios implemented this bank’s cross-sell campaign strategy across 20+ campaigns and six channels using a mathematical campaign optimization routine. We mapped the bank’s existing contact strategy and campaign selection rules, as well as their analytic scoring routines and back-end performance measurement strategy, to the process required to produce their campaigns in production. We completed the implementation in less than four months.||For the campaigns launched in the first two months of production, the performance measurement results, compared to the same campaigns year-over-year, yielded a gross sales rate of $160 per contacted customer, compared to prior year sales rate of $138. In that two month period, this campaign revenue to the bank represented an incremental $7 million gain over the prior year’s results.||Harmony|
|40||Banking||Retail bank||Compliance and Governance||Customer marketing model lifecycle management||The model development and model validation groups of this bank needed assistance in modernizing their analytic model processes in order to meet and exceed global guidelines on model transparency and auditability.||We developed business processes for this client to deploy and track model assets, track team workflows on model creation and review, track the ongoing mathematical performance of their models, and to support model transparency and governance goals for the underwriting function. We imported five models into the tracking tool and we trained the team on how to import the remainder of their model portfolio.||Our client substantially reduced the time they allocated to model tracking, audit support and documentation, so they could refocus more time on the creation of new and updated model assets.||Metronome|
|41||Investments||Brokerage||Analytics modernization||In-database model development and deployment conversion||Corios had already helped this retail brokerage to implement a mathematically optimal leads distribution engine to allocate the right offer to each financial advisor ("FA") and client. However, the scoring platform used by this engine could only refresh every 72 hours, so our client struggled to deliver fresh daily leads to their FAs.||Corios converted our client's model development and deployment processes from a classical SAS+Oracle architecture into an in-database strategy with SAS running alongside an EMC GreenPlum data appliance. We developed modern business processes for model development and deployment. We also developed a custom routine for the analytics users that translated their standard model development strategy into a 100% in-database approach, ensuring rapid deployment.||Corios' in-database model development and scoring platform and processes reduced our client's work cycles from months to weeks, and reduced model scoring run times for a batch of 10+ million customer records from several days to a few hours.||Rosetta|
|42||Credit issuer||Credit card issuer||Revenue forecasting||Acquisition profitability forecasting and offer optimization||The credit product management team of this subprime credit issuer needed to be able to project the profitability and default exposure of customer vintages that were less than 12 months old. However, their tools for performance forecasting were spreadsheet-based which meant that scenario development required days, and changes to scenario structures required weeks of manual effort.||Corios developed a suite of forecasting models and a modernized analytics process for our client. We also developed a card offer optimization model that gave their analysts real-time ability to design new card acquisition offers, examine the likely profit and risk exposure of such a model, and compare 60-month forward performance curves to business as usual.||Corios' strategy reduced the time required to build a new forecast scenario from six hours down to five minutes; this enabled the analysts be able to rapidly develop multiple scenarios and stress tests on forecasted credit risk. This will increase the future profitability of all their offers by calibrating product strategy to customer response, utilization and risk behavior.||Forte|
|43||Banking||Retail bank||Risk mitigation||Modernize commercial lending judgmental scorecards||The commercial credit risk and underwriting team for this bank had developed judgmental scorecards as spreadsheets that were emailed to lending relationship managers. The operational and financial risk exposures of this approach were substantial, and the bank wanted to centralize and secure the entire platform.||Corios developed a process for the commercial lending risk management team to modernize their judgmental scorecards. We converted their spreadsheet-based process into a modern approach that separated the mathematical equations used in lending strategy from the user interface used in the branches and mobile applications for customer loan application conversations.||Corios' strategy enabled the bank to track their commercial lending performance on a more reliable, transparent and governed basis. We translated 8 commercial lending scorecards into test version, which the bank now has in production at all their client touchpoints.||Tempo|
|44||Retail||Retailer||Revenue growth||Promotional response and store-level analysis||A nationwide office products retailer needed to overhaul their business paper products promotional strategy. The client couldn't attribute their market mix and trade promotional strategies to in-store sales or determine the best promotional timing relative to baseline sales volume.||Corios quantified the sales impacts associated with online, newspaper, in-store, and competitor paper promotions for our client. We segmented the data based upon store size, and we suggested which promotional characteristics to pursue more aggressively and which to avoid.||Using Corios' strategy, our client made better use of their advertising budget by improving promotional key wording and cross-promotion tactics, timing the promotional release day of the week most favorably, and better understanding the trade-off sales interactions among promoted and non-promoted products.||Tempo|
|45||Energy||Electric utility||Load forecasting||Peak load and demand forecasting system||A large electric utility was struggling with their short term and long term price and demand forecasting, since they relied on diverse and disconnected software applications to complete their work. Individual sets of reports had to be manually cobbled together to form a consolidated, calibrated forecast.||Corios built a tailored software platform for price and demand forecasting and optimization that was used all the client's team members. This platform tied in all of their existing data, and supported their current and future data and analytics requirements.||The price and demand forecasts yielded by this system helped the client to improve the accuracy of their forecasts, complete the calibration of short and long term forecasts in a single step, and enable the team to focus more time on insights generation and less on manual data manipulation effort.||Lightning|
|46||Energy||Electric utility||Load forecasting||Evaluation of smart grid impacts||An electric utility had implemented a smart grid so that they could start monitoring the effects of household energy technologies on consumer demand. However, they struggled with how to perform this measurement in comparison to baseline demand for electric power by these customers.||Corios developed econometric models and assessed the impact of enabling technologies and dynamic rate plans on customers’ energy consumption for that utility’s smart grid pilot program. Advanced metering infrastructure (AMI) data from the utility’s pilot program were analyzed using SAS software that integrated with the utility’s Teradata data warehouse.||The Corios strategy and analytic insights identified the effects of different enabling technologies on customer power demand, including programmable communicating thermostats, web portals, and in home displays, as well as the effects of Time of Use and Variable Peak Pricing rate plans.||Lightning|
|47||Energy||Electric utility||Load forecasting||Demand and load forecasting||An electric utility needed to develop revenue and demand forecasts for their long-term (20 year horizon) integrated resource plan.||Corios developed monthly econometric sales forecasts which incorporated the effects of macroeconomic drivers, weather, retail electric rates, and other seasonal drivers to predict long-term (20 year) retail sales.||The forecasts that Corios produced for our client substantially outperformed their de facto forecasts and were judged as suitable for their long-term resource planning.||Lightning|
|48||Energy||Electric utility||Load forecasting||Demand and load forecasting||Our client is a regional electric utility serving the southwestern US. Their incumbent load forecasting application was built at the county level, and they needed to make these forecasts more granular for grid modernization purposes. They also needed their forecasts to tie top-down and bottom-up results, and they needed to improve the way they handled future weather scenarios.||Corios developed a system for programatically producing hundreds of zip-code level forecasts of monthly electricity sales. A mix of econometric and time series models were assessed for a variety of model-fit criteria before those candidate models were accepted. Reconciliation of the disaggregated forecasts with the retail system load forecast was then performed to ensure consistency between the forecasts.||The client successfully deployed the new forecasting engine and used the updated forecasts to generate more accurate and actionable forecasts in their grid modernization process.||Lightning|
|49||Credit issuer||Retail bank||Compliance and Governance||Analytic model validation and performance tracking||The model development and model validation teams of this retail bank needed assistance in modernizing their analytic model processes in order to meet and exceed federal guidelines on model transparency and auditability.||Corios developed business processes for our client to deploy and track model assets, track team workflows on model creation and review, track the ongoing mathematical performance of their models, and to support model transparency and governance goals for the underwriting function. We also trained the team in how to import new models into the tool and how to introduce new users to the process as they join the team.||Our client substantially reduced the time they allocated to model tracking, audit support and documentation, so they could refocus more time on the creation of new and updated model assets.||Metronome|
|50||Credit union||Credit union||Customer profitability||Campaign management||The marketing division of this credit union had been wholly dependent on a marketing agency to run their campaigns, which as a result tended to be plain vanilla and treated most members as if their relationship with the credit unions were all equivalent. The credit union wanted to take advantage of their member data warehouse in segmenting and targeting campaigns, and in attributing responses to those campaigns in future efforts.||Corios automated the production of marketing campaigns for our client using their enterprise data warehouse as the primary data source augmented with data tables supplied by external data source providers, such as Experian. We replicated many of their campaigns in the new platform, trained them how to replicate and build new campaigns, how to modernize their campaigns using the new tool, and taught them how to use test-and-learn and analytics in their targeting strategy.||This increased credit union usage by members, strengthen and deepen member access, provided real-time access to data, and increased cross-sell & retention.||Harmony|
|51||Insurance||Insurance carrier||Analytics modernization||Analytic model deployment platform||A nationwide property and casualty insurance provider was suffering from claims scoring processes that required more than 5 minutes to score per claim, not to mention model deployment cycles of 12 to 18 months per release. The most significant challenges were that they recoded every analytic model from SAS into C, and that they merged over 30 data sources to score most types of claims.||Corios developed an enterprise-wide analytic risk scoring data platform and developed a real-time risk scoring process for our client. We refactored actuarial model code assets for loss reserving, subrogation, and fraud. We created a modern actuarial predictive model deployment system with version control, change management, model development, model deployment, model scoring in batch and real time, and model performance tracking.||The new Corios platform reduced by more than 50% the analytic model deployment time, and reduced scoring time per claim by 70%. The client was so pleased with the first major project on their property claims portfolio, that they hired us again to convert their bodily injury portfolio as well.||Rosetta|
|52||Manufacturing||Vehicle manufacturer||Cost reduction||Telematics and claims analytics platform||An international truck manufacturer set a strategic global objective to surpass their competitors on vehicle reliability. They had truly massive data repositories but lacked the ability to harness it, make it predictive, and take action on the trucks most at risk of technical or mechanical failure.||Corios built an analytics platform to predictively identify vehicles at risk of mechanical problems before they happen, by combining real time data from telematics alerts, service center geofencing, repair histories and travel and road conditions. The predictive analytics models incorporated breakdown, transit and arrival time, diagnosis, parts, shipping, repair, notification and departure phases of the service chain. Using survival and transactional analyses, we identified factors associated with increased sensor failures and categorized vehicles by risk.||Corios helped our client increase the North American fleet uptime from its current state of 96% towards it targeted state of 99% through predictive root cause analysis on the dealer-service chain for hundreds of thousands of vehicles and half a billion telematics alerts.||Tempo|
|53||Energy||Electric utility||Load forecasting||Demand and load forecasting||Our client was an electric utility based in the US Midwest who serves over 2 million customers. They needed a custom electric load forecasting solution for the multiple states in their service territory.||Corios built a forecasting platform for our client that modernized forecasting processes including manually pre-forecast data preparation, statistical modeling, and post-forecast reporting.||The solution integrated with the utility’s meter data management systems to extract data for forecasting models, stored forecasting results, and summarized forecast results for management-level reporting. This enabled our client to meet their regulatory reporting standards successfully.||Lightning|
|54||Energy||Electric utility||Load forecasting||Peak demand forecasts||Our client is an electric utility serving millions of customers in the Southern US. They needed a peak load forecasting application to assist them with system planning and grid modernization.||Corios built a forecasting platform for our client that replaced an existing industry-standard software package and integrated with the utility’s long term energy forecasting system to predict and visualize hourly electricity demand based on weather conditions and seasonal trends.||The solution integrated with the utility’s meter data management systems to extract data for forecasting models, stored forecasting results, and summarized forecast results for management-level reporting. This enabled our client to meet their regulatory reporting standards successfully.||Lightning|
|55||Insurance||Health care payor||Customer profitability||Customer profitability segmentation and acquisitions campaigns||A major health care payor organization struggled with their mid-market portfolio, which was unprofitable nearly across the board. However, the payor could not identify which groups and brokers contributed to profitable business, so they were effectively blind for how to resolve this situation.||Corios developed a segmentation strategy for our client to identify profitable and non-profitable small business sector groups, the brokers that call on them, and the relationship manager strategies for investing in brokers that would restore profitable growth in that sector. The segmentation strategy evaluated income drivers (premiums), cost drivers (utilization) and sales drivers (prior sales and marketing investments with groups and brokers), and recommended an actionable path forward to re-direct broker activities towards more profitable growth.||Using this platform, Corios and our client designed an acquisition campaign that maximized likely profitable growth utilizing the largest territory in that payor organization's footprint, using internal and third party information regarding groups not currently under coverage.||Harmony|
|56||Energy||Electric utility||Cost reduction||Workforce capacity planning||One of the country's largest electric utilities suffered rapidly increasing costs and poor service levels to respond to localized customer service outages. Their demand forecasting and scheduling routines were largely manual, and a recent investment for real time truck routing wasn't addressing the problem, since the staffing for trucks and regions had not been addressed.||Corios created an analytics platform that forecasted work load, employee availability, allocates weekly FTE to forecasted workload via a mathematical optimization algorithm.||Corios' strategy saved the utility considerable cost by allocating the ideal staffing to the forecasted workload, upwards of $6 million per year.||Lightning|
|57||Banking||Commercial bank||Customer profitability||Marketing analytics platform and campaign optimization||A top 20 American bank had no platform to track their omnichannel marketing communications and resulting customer behavior. They also had no ability to view the customer relationship across all portfolios (retail, business banking and wealth management) or to identify their most valuable customers.||Corios developed an enterprise-wide marketing analytics platform, consisting of consolidation across more than 20 systems of record, householding of customer relationships, campaign management, integration with their major outbound channels, closed loop attribution of offer effectiveness, a predictive model factory (with more than 70 new models) and a next best offer engine, over just a 36 month period of time.||Through Corios' strategy and implementation, the client now has a world-class marketing platform, has launched closed-loop campaigns generating millions of dollars of new income, and is poised to expand this platform into their digital-centric channels and provide real-time leads into their enterprise CRM platform.||Harmony|
|58||Banking||Retail bank||Customer profitability||Next Best Offer campaign optimization||A top 20 American bank needed to determine how to arbitrate among retail banking offers and channels, as well as avoid customer fatigue, and grow income from cross-sell campaigns.||Corios developed a retail banking next best offer engine for more than 10 products and 3 channels. This engine consists of predictive response and revenue models, and a mathematical optimization engine to maximize income generation while balancing customer contact patterns.||The client's retail offers generated millions of dollars of new revenue in the first 12 months of operation, more than paying for the investment in less than a year.||Harmony|
|59||Banking||Commercial bank||Customer profitability||Next Best Offer campaign optimization||A top 20 American bank needed to determine how to arbitrate among business banking offers for treasury management, merchant services and commercial credit cards, and distribute these leads to business banking relationship managers for large account cross-selling.||Corios developed a business banking next best offer engine for more than 25 products. This engine consists of 50 predictive response and revenue models, and a mathematical optimization engine to maximize income generation while balancing customer contact patterns, relationship manager calling capacity, and offer rotation. We also integrated the next best offer engine with the bank's enterprise CRM platform to enable closed-loop tracking of lead disposition.||Corios developed and rolled out an initial release of the client's business banking next best offer engine in less than 8 weeks. We raised response rates from 1.8% (control) to 2.9% (campaign) and increased fee income per client by 57%.||Harmony|
|60||Retail||Retailer||Customer profitability||Digital marketing campaign optimization||A major retailer's digital division, who sends out tens of millions of promotional emails per week wanted to start using analytics to determine how to be more selective in sending offers to customers, in order to both drive income growth, as well as avoid customer fatigue and to meet their agreements with their merchandisers. However, their initial attempts at using this approach were not successful.||Corios identified that the analytics scoring team and CRM team needed to develop a mutual understanding of how the platform should work. We developed new offer strategies and customer contact patterns, and made recommendations about enhancing the analytic model scoring approaches for seasonal, holiday and channel effectiveness.||Corios' strategies collectively produced a sales revenue lift of 20-25% through the use of the marketing optimziation platform for the digital promotions of this retailer client.||Harmony|
|61||Banking||Retail bank||Revenue growth||Mortgage offer timing optimization||A major American bank wanted to improve their mortgage-generated income by finding customer opportunities in a more timely basis. They had developed a massive new data repository of financial transactions and omnichannel interactions, but hadn't yet harnessed this to predict the moment of truth, when their customers were considering moving their home.||Corios first developed some innovative time-to-event and event sequence analytics that significantly increased model predictive performance for the mortgage acquisition sales event. We also reduced the model refresh cadence from monthly to daily. Next, Corios deployed the data and the models into a Hadoop data lake, which allowed us to score every customer every day and reduced the scoring job time from 50 hours to 5 hours.||With the improvement in mortgage sales opportunity identification, the bank stands to gain more than $10 million annually, and has more useful behavioral information to target different types of customers. Corios is now extending this solution to cover other retail products like deposits, credit cards, home equity and other lending.||Veloce|
|62||Banking||Retail bank||Customer profitability||Campaign platform build and implementation||A large American bank had developed their marketing platform over a decade but badly needed to modernize their platform. Their processes were largely manual and work-intensive, with very little documentation and limited ability to support any audits in support of regulatory review of lending campaigns.||Corios implemented a modern campaign management application, helped migrate the client's old customer database to a new platform, and modernized over 600 campaigns into the new platform.||Corios' implementation enabled the bank to reduce 600 manually-implemented campaigns into less than 200 automated campaigns, streamlined their quality assurance for each campaign, integrated analytics and scoring into the platform, and enabled the transformation of the campaign team from a production line into a solution engineering mindset.||Harmony|
|63||Banking||Retail bank||Customer profitability||Cross sell marketing campaign optimization||One of the largest banks in North America needed to automate their campaign optimization processes for more than 30 campaigns over 6 channels on a daily basis.||Corios developed a fully automated platform for the bank to execute all their campaign optimization processes on a daily basis, with scheduling, triggered execution and campaign management integration.||Corios completed this work in less than one month, and fully automated all these tasks so that the marketing analytics team could focus on strategy and insights rather than on mechanics.||Harmony|
|64||Banking||Retail bank||Customer profitability||Mortgage and HELOC campaign segmentation, targeting and optimization||This retail bank wanted to modernize their direct marketing capabilities by building a customer-centric database and enabling targeted marketing offers through multiple channels, but didn't have the requisite skills or architecture experience.||Corios helped our client to design a customer analytics architecture that would support pre-approved lending offers, with a focus on mortgages and home equity lines of credit. We assembled their data, overlaid credit bureau feeds, built targeted marketing analytic scoring models for acquisition, cross-sell and refinance offers, and developed several campaigns.||We built the entire customer analytics architecture in less than six months, generated substantial predictive lift on their mortgage and home equity campaigns relative to business as usual, and established buy-in with the compliance and regulatory team in terms of the campaign and targeting analytics strategy.||Harmony|
|65||Energy||Electric utility||Risk mitigation||Anti-money laundering case management system||This large electric power utility needed to improve their capabilities for managing alerts and cases tied to anti-money laundering threats.||Corios implemented an anti-money laundering alert and case management application for the utility client.||Corios completed the implementation and customization of this platform in under 12 months and satisfied all the client's requirements.||Tempo|
|66||Banking||Retail bank||Compliance and Governance||Analytic model repository strategy and coaching||This rapidly growing retail bank needed to implement an analytic model repository in conjunction with federal regulatory and compliance requirements.||Corios developed business processes for our client to deploy and track model assets, track team workflows on model creation and review, track the ongoing mathematical performance of their models, and to support model transparency and governance goals for the underwriting function. We also trained the team in how to import new models into the tool and how to introduce new users to the process as they join the team.||Metronome|
|67||Credit union||Credit union service organization||Analytics modernization||Marketing campaign database optimization||Our client had developed a large customer database containing more than 10 million active credit card accounts and billions of records of detailed transactions. Their customer analytics required a 26 hour cycle time to complete when new customer records were added to the repository, making it very challenging to use these analytics in any of their customer retention campaigns.||Corios redesigned the customer data platform to embed all the analytic functions inside a distributed data appliance that was already supporting the enterprise data warehouse. We also designed and implemented a solution to align customer identities across multiple cards, so that even when a customer was issued a new card, the longitudinal link back to the same customer’s transaction history was maintained.||Through Corios’ redesign and the collaborative delivery work to implement this solution in production, we converted the 26-hour customer analytics process into one that runs in 10 seconds. This opens up opportunities for our clients’ marketing analysts to use the power of customer value over the entire lifecycle of the customer to identify opportunities for new campaigns.||Rosetta|
|68||Investments||Wealth management||Customer profitability||Customer lifetime value measurement and optimization||An institutional investments client had developed a client lifetime valuation model but struggled to use it to predict the impacts of pivoting business initiatives, changing relationships, or simulating new business opportunities is impossible, which were essential to pricing the client relationship. The CLV model lived in a complex spreadsheet, making it unwieldy to scale to the entire business.||Corios centralized and scaled the CLV model, and placed it in a web portal that offered investment portfolio owners to run what-if analyses that maximized the value of their client services and lifetime value.||Our client could access client lifetime values from a broad number of devices, including on the road, and recalculate CLV scenarios for each institutional client on the fly, enabling them to be far more prepared for negotiating each client's book of business.||Forte|
|69||Banking||Retail bank||Risk mitigation||Fraud model validation services||This retail bank needed to meet regulatory audit requirements for three fraud models, including credit transactions, debit transactions, and online banking wire transfers.||Corios developed model validation requirements, replicated each of the fraud models on independent platforms, and documented the findings for each model. Corios also developed operational recommendations to assist the bank in being better prepared for future audits.||Corios delivered actionable recommendations to the bank related to skills and tool development, internal rule and model testing processes, develop a broader analytic basis for model development and validation, quantitative measures of model performance, change management and version tracking, and fraud model documentation.||Tempo|
|70||Credit union||Credit union||Customer profitability||Member database build, householding and campaign optimization||A credit union had no platform to track their omnichannel marketing communications and resulting member behavior. They also had no ability to view the customer relationship across all portfolios (retail, business banking and wealth management) or to identify their most valuable customers.||Corios developed an enterprise-wide marketing analytics platform, consisting of consolidation across more than 6 systems of record, householding of member relationships, campaign management, integration with their major outbound channels, closed loop attribution of offer effectiveness, and an analytic model factory in under 8 months.||Through Corios' strategy and implementation, the client now has a world-class marketing platform, and is poised to expand this platform into their digital-centric channels and provide real-time leads into their enterprise CRM platform.||Harmony|
|71||Energy||Electric utility||Load forecasting||Hourly load forecast for every electric feeder||In order to respond to market demand for implementing DERs, California regulatory expectations for DER adoption and greenhouse gas emissions, and opportunities to improve the safety and resiliency of the distribution system, Southern California Edison needed a Long Term Planning Tool (LTPT) platform for forecasting hourly MW demands for every one of its ~7000 distribution assets down to the circuit level, in order to achieve 3 strategic objectives: 1. Facilitate a consistent and integrated load, DER and generation forecast; 2. Facilitate a risk-based, integrated system analysis across the organization; 3. Facilitate the development of optimal grid solutions, including evaluating DERs and traditional infrastructure problems||Corios designed and implemented the LTPT platform for SCE which produces 8760 load curves for every substation and circuit, for hundreds of planning scenarios and weather forecast types. LTPT’s econometric and time-series forecasting capabilities are based on several components, including long-term monthly energy forecasts where usage is modeled as a function of economic drivers, weather, and calendar effects, hourly demand models that estimate load profiles as a function of weather and calendar effects, and the extrapolation of DER and LGP forecasts to the hourly level using a variety of load profiles. The final output is a ten-year forecast of hourly load, disaggregated by source (base, DERs, and LGPS) for every asset in SCE’s service territory.||SCE can now forecast in advance the likely frequency and duration of capacity violations, and estimate the peak shifting that occurs due to DER adoption, estimates both energy and demand consumption, and calibrates the entire forecasting system to the California Energy Commission top-down forecast for the SCE territory.||Lightning|
|72||Manufacturing||Vehicle manufacturer||Risk mitigation||Warranty claims decisioning engine||Our client, an international truck manufacturer processes 10,000-15,000 warranty claims a week. They have the ability to automatically pay the claim, or to set it aside for a manual review. They needed to reduce warranty payout, and pay claims faster, with fewer people and time available per claim.||Corios built a modularized rule framework such that analysts can develop and promote highly customized rules to target claim populations at a lower granulairy than the previous solution. By leveraging this manufacturer's historical claim, product, coverage, labor code, included material,case, and telamatics data, Corios was able to develop over 200 rules that idenitfy the 10% of incoming claims that violate, or are highly likley to violate, warranty policies.||Our client now pays claims 15% faster, increasing customer satisfaction, while reducing labor payout by 12%. This solution is directly responsible for 70% of all claims adjustments made by the client.||Tempo|
|73||Manufacturing||Vehicle manufacturer||Cost reduction||Purchased coverage profitability analytics||Our vehicle manufacturing client manages a portfolio of extended service and maintenance contracts to help customers reduce operating costs and increase vehicle uptime. Our client's incumbent data analytics solution is used to define content of offering, set standard costs and commercial pricing to optimize the performance of the portfolio. Incumbent solution was slower than desired, code-heavy, and difficult to scale across the entire portfolio at the level of detail desired by the key users and stakeholders.||Corios developed a pilot project for our client on AWS to visualize the price and coverage term as well as impact of hundreds of thousands of scenarios for its contracts. Compared to the incumbent approach, the pilot solution reduced the amount of code by 50%, ran the pricing scenarios in memory, reducing the calculation time for each scenario from 4+ minutes to real-time (e.g., virtually instantaneous).||The ability to analyze data real-time enabled our client's leadership to work more efficiently, optimize pricing and reduce claim costs to materially boost profitability of their portfolio.||Tempo|
|74||Credit union||Credit union||Customer profitability||Member attrition analytics||This large credit union runs a membership loyalty program, and the CMO wanted to understand the drivers of attrition so they could maintain member activity. The credit union team couldn’t project program attrition, the time of onset or what to do about it.||Corios developed predictive scoring routines to identify the leading factors of loyalty program attrition for each of three behavioral segments of members, and the specific types of change in member behavior, so that the appropriate action could be taken to save the member's status in the program before they left for good.||Corios' models achieved lift of 2-3X over control group, identified attriters two months ahead of the actual event, and identified the behaviors that members exhibited ahead of the attrition event so that proactive outreach could be micro-targeted.||Harmony|
|75||Credit union||Credit union||Customer profitability||Business simulation gamification platform||One of our clients, a large credit union in Southern California, asked me to speak at their quarterly executive offsite, with their top 100 executives. I proposed instead, let’s do something fun and unexpected—let’s play a game! We designed a business simulation game for the execs to play. This game was a blend of game theory, teamwork, AI/machine learning, and good old fashioned competitive spirit.||Corios built a banking simulation game, where 10 teams (of 10 executives each) play the role of running the credit union for a year, dividing into 3 turns. Each turn, each team simultaneously decides how to invest 45 investment points into advertising, cross-sell, retention, customer care and credit risk strategies. Rewards were given out at the end of the game for the team who made the most money, who saved the most customers-at-risk, who sold the most products, and who made the highest Internal Rate of Return on their investment.||The team-building exercise was a complete success. “I only wish we could play again and spend even more time doing it!” – Chief Executive Officer “Thanks for all you did to make the Management Team activity awesome!” – Chief Marketing Officer “One of the most exciting, fast phased, fully engaged collaboration team events we've ever done in Management Team. I would do it again as well.” – Director of Digital Channels||Harmony|
|76||Banking||Commercial bank||Customer profitability||Treasury demand development||Our commercial bank client wanted to increase the client uptake of treasury management services, corporate credits card and merchant services. Their own cross-sell rates were behind industry averages, their commercial bankers did not have visibility into which treasury services their clients might need, or even how to start the conversation.||Corios build a predictive model factory of 80 models on 40 treasury services (i.e., adoption rate and predicted income generation) to select the ideal 3-4 treasury services for each commercial banking customer. Corios also built a closed-loop integration with the bank's CRM to deliver, monitor and refresh these leads based on execution and feedback from bankers and customers.||The adoption rate lift in treasury management close rates climbed from 2:1 to 7:1 depending on products, compared to the control group. On average, the typical close rate jumped from 5% to nearly 20%.||Harmony|
|77||Banking||Retail bank||Compliance and Governance||CECL compliance||A regional retail bank needed to fullfil new requirements for CECL regulatory accounting standards. The bank needed not only to implement new loss forecast models in a controlled manner, but also needed to implement on a robust platform that would enable their growth-oriented corporate strategy and integrate easily with additional credit reporting needs.||Corios implemented a set of model groups, controls, reports, data generation, and data quality processes on the SAS Expected Credit Loss (ECL) platform. Corios also developed a security model and set of controls to assure components of the CECL modeling and allowance processes are in compliance with all regulatory standards.||The platform has proven reliable over the last two years. During which time, the bank began the process of completing a merger of equals with another regional retail bank. The platform that Corios built has been chosen as the go-to platform for all CECL modeling activities and is in the process of expanding to meet the newly formed bank's needs.||Forte|
|78||Banking||Retail bank||Compliance and Governance||CECL compliance||A regional retail bank needed to fullfil new requirements for CECL regulatory accounting standards. The bank needed loss forecast models in a controlled manner, and enable their unique way of incorporating customized macro-economic forecasts to inform expected loss modeling.||Corios implemented CECL models in the SAS Expected Credit Loss (ECL) platform for a robust, controlled solution to meet CECL regulatory requirements. To meet unique macro economic forecasting needs, Corios integrated an automated Moody's SFTP process that specifies forecasts to prepare, reports on characteristics, loads data into the solution, and audits all actions related to the macroeconomic data.||The Client is enabled with robust credit risk platform that includes a custom data preparation module that meets both analytic and accounting needs, giving the client confidence and transparency for their loss forecasting processes.||Forte|
|79||Banking||Retail bank||Customer profitability||Marketing mix modeling strategy||A Canadian bank sought to increase its US retail presence by acquiring a number of regional banks along the US East coast. The bank needed to balance investments in brand development with local market sales support. Corios was contracted to design a marketing mix modeling environment that would quantitively balance the competing investment needs between brand investments, new market entry investments and dominant market management investments across all marketing spend.||Corios designed a set of analytics and the supporting data collection and management architecture that would allow for the measurement of individual marketing investments by market across time. This platform also supported robust forecasting at the local market level that allowrd marketers to predict the economic return on each type of marketing investment for each local market.||The client has a robust analytical environment to plan and measure all marketing investments. Further, every data element and analytic model is fully documented so that no internal disputes develop due to 'black box' predictions.||Harmony|
|80||Banking||Brokerage||Customer profitability||Marketing Optimization||Our client, a retail brokerage firm, wanted to implement optimized marketing business rules for their trader enrollment program.||Corios designed, implemented, and enabled the client's marketing analytics team on a set of propensity model driven business rules. Corios also advised on measurement and reporting techniques to compare optimized and un-optimized campaigns||Rigorous testing showed optimized campaigns generated significantly more customer engagement in their trading and training platforms, resulting in higher retention and revenue rates.||Harmony|
|81||Banking||Corporate bank||Revenue forecasting||Price optimization||Sales and market competition put downward pressure on product and services pricing. The fixed costs of a 'premium brand strategy' could not be recovered until substantial enrollment levels were reached for any given association relationship. Our client had no mechanism to measure these costs or monitor enrollment thresholds which led to multiple unprofitable association relationships.||Corios developed a cost measurement platform that allowed our client to understand the cost for any service provided to any association at any enrollment level. These tools enabled the sales teams to structure new relationships such that profitability was achieved sooner, while allowing the operations teams to focus enrollment efforts on underperforming association relationships.||Our client achieved several positive outcomes from this engagement. These included: reduced organizational stress between sales and operations; fewer new relationships being onboarded with low probability of profitability; and improved enrollment performance for low-profit associations.||Forte|
|82||Banking||Insurance carrier||Loss forecasting||IFRS17 compliance||Our client manages a corporate insurance portfolio, and their actuaries needed to implement the IFRS17 standard. The purpose for this engagement is to better estimate the Value at Risk (VaR) on their portfolios' liability for remaining coverage (LRC) and liability for incurred claims (LIC).||Corios' challenge was to transform our clients' actuarial science approach, originally designed using spreadsheet and Python-level analysis, and scale it to a) a more granular level: expanding from portfolio and tranche-level analytics down to handling hundreds of thousands of individual insurance contracts, and b) high-intensity Monte Carlo approaches for simulating a wide range of potential risks (aka stochastic simulation).||Corios helped our client reduce run-time from 3 hours per contract, down to 3 hours for hundreds of thousands of contracts and hundreds of thousands of stochastic simulation trials. We implemented this cash flow modeling engine across 10 insurance portfolios and helped our client meet the regulatory deadline for IFRS17.||Forte|
|83||Insurance||Insurance carrier||Analytics modernization||Analytics platform migration||Our property and casualty insurance client had 800 users on a legacy analytics platform, and wanted to move all their users, data and workloads to a modern cloud analytics platform using open source capabilities.||Corios used our Rosetta software/sevices methodology to inventory the workloads and data of all users, to move the workloads upstream to the databases from where the data originated, or downstream into the cloud platform, depending on the use case and required functionality. We also translated all the active legacy workloads into either SQL, Python or Spark.||Our client reallocated 150TB of data storage, saving ~$0.5M annually. Right-sizing their software licenses saved another $800K annually, and consolidating duplicate processes and staff led to $4M in expenses savings annually.||Rosetta|
|84||Credit union||Credit union||Compliance and Governance||CECL compliance||Our client (a large east coast credit union) was having a difficult time understanding and implementing data mart development, data transformation, and delivering value from a newly built data lake. The client selected a hosted platform on which to run its CECL, AML, and other Credit Risk processes, and this high-visibility project had no room for delays.||Corios used our Tempo methodology and credit risk expertise to design, implement, and enable project management, requirements gathering, development, and validation processes.||The client's data preparation is automated, on-time, complete, and accurate, demonstrating return on the clients investment into the cloud-based data lake for their credit risk projection purposes. Corios also performed a gap analysis, highlighting data quality issues that need to be fixed in the source system, and are working with the client to implement a new and improved data governance strategy to improve future forecasting efforts.||Forte|
|85||Credit issuer||Credit card issuer||Revenue growth||Attrition risk modeling||Our credit card issuer client was dealing with a cascading challenge. As a service provider to multiple lenders, our client relied on each lenders' customers to transact with the cards it issued. If the member chose to use an alternative card, both the lender and our client were negatively impacted.||Corios developed a set of analytics that would identify individual credit union members at risk of migrating to an alternative card for the majority of their transactions. These analytics were able to identify the individuals, recommend intervention strategies, and deliver these messages to the credit union for action.||Our client deployed these models to reduce attrition and developed a >25% reduction in card substitution, and a 5% growth in revolving balances.||Tempo|
|86||Retail||Retailer||Analytics modernization||Digital analytics roadmap||RunAMZ helps premium brands win in the world’s biggest and fastest moving marketplace, Amazon.com. RunAMZ, based in Portland, Oregon, helps their clients manage every aspect of their Amazon business, from marketing and page optimization to MAP enforcement and logistics. As the RunAMZ business has grown, company management found they needed to improve the timeliness and relevance of operational, marketing and promotional metrics to help them measure, anticipate and optimize their clients’ business performance.||RunAMZ consulted with Robin Way, President of Corios, on a strategy to define a strategy for management decision metrics and business performance optimization. This strategy will detect the drivers of performance from their historical activity, and anticipate the best decision alternatives for RunAMZ managers to adjust their client’s business, as those business drivers evolve.||RunAMZ now has a vetted strategy for business performance monitoring and forecasting that they can prioritize, implement and measure as their business grows. This strategy can be implemented in a modular fashion over time, using Amazon Web Services capabilities for big data and analytics, augmented by the business analytics competencies that Corios brings to the table.||Rosetta|
|87||Banking||Credit card issuer||Risk mitigation||Fraud prediction modeling||Our client needed to enhance its first-party fraud detection for both account takeover and transaction fraud use cases. Its previous predictive models were several years old and didn't perform sufficiently. Our client also wanted to compare how these models could be built and perform on open source frameworks in addition to their traditional analytics platform.||Corios developed account- and transaction-level attributes for cardholders, and developed over 50 candidate models in SAS, finally selecting the champion models which substantially outperformed the incumbent models by several K-S points. We also translated these models into PySpark, running them on the Amazon EMR platform, and taught the client how to compare and contrast the two approaches for model construction and validation.||The champion models were put into production by our client on their credit bureau platform, and their performance in the field has successfully met the client's expectations. The client is active in building their analytics on both their traditional platform and is experimenting with open source approaches.||Tempo|
|88||Insurance||Health care||Analytics modernization||Revenue integrity workload modernization||Our client had developed a revenue integrity analytics workload that required 10 staff and 4 months per year to execute. The run time for the application often required weeks to run, usually required manual restarts, and the client was concerned they were missing crucial mistakes in regulatory filings for health care claims.||Corios migrated the revenue integrity analytics workload to Amazon Glue and converted all the code to PySpark. Corios provided the lead role on data engineering, design and implementation and assisted with testing and validation.||The modernized analytics application saved the client 4 FTE-months of productivity per year, reduced data processing time by 98%, improved annual time to market by 12 weeks, and improved the data integrity of regulatory filings by over 30%.||Rosetta|