Insurance Innovation Award-Optalix
0 likes121 views
The Modelling API was developed to quickly convert client models into accessible APIs. It allows users to instantly convert Excel and Python models into hosted APIs, input external data for models, and send outputs to a management system. The API enables rapid model deployment and updates, reducing delays between insights and system updates. Experience monitoring dashboards track model accuracy in real time. Major insurers and banks use the Modelling API and Optalitix portals to gain data-driven insights.
Insurance Innovation Award-Optalix
- 1. Modelling API Modelling API May 2018
- 2. We give systems a brain 2 We provide tools for insurers & banks to use data analytics and machine learning Our sales platform converts client leads more intelligently, reducing acquisition costs Our retention platform uses predictive analytics to reduce lapses Our claims system identifies opportunities to reduce claim costs We enable model deployments into API’s accessible by other systems
- 3. The Modelling API converts any model into an accessible API 3 CONVERT RETAIN Sell to customers Smart retention strategies CLAIM Intelligent claims handling UNDERWRITE Intelligent underwriting MODELLING API Modelling Tools Enable AI in existing systems Configurable Portals
- 4. Optalitix Overview 4 The Modelling API was developed to meet implement client models quickly. We have now adapted it for external use. It is used by a number of clients for models ranging from credit affordability checks, product recommendations, sales and retention propensity measurement and even petrol pump shortfalls. The functionality is constantly being enhanced based on specific client needs. It now allows user to: • Instantly convert Excel and Python models into system hosted API’s • Host the API and produce documentation for external systems to connect • Input data for models from external sources(e.g. external databases like Experian) • Send model outputs into a model management system to track outcomes and use them operationally. Models Decision engine API External Systems
- 5. How it works 5 Decision engine Testing ScreensReplicated Model Optalitix Experience monitoring Data Store Modeling API API Websites Admin Systems Mobile Apps CRM systems Modelling Team Insurer & Broker Systems Step 1: Model uploaded and replicated in decision engine Step 2: Online model testing screens generated by the system Step 3: API set- up using system generated SDK documents Step 4: Database stores API data Step 5: xperience in real time Testing Screens
- 6. Delays in model implementations reduce model accuracy 6 Initial Model Build Experience Monitoring Model Update Model conversion into system System Update Experience Monitoring To shorten the cycle, the systems supporting the models need to be optimised 1-2 months 3-6 months 2-3 months 1-2 months 1-3 months 2-4 months 2-3 months Business insight is inaccurate due to delays 10 – 20 months
- 7. Optalitix’s Modelling API ensure rapid data insight 7 Instant system build from existing models System integration is easy Experience monitoring dashboards Data for model updates is always available Model updates require no system intervention Instant model conversion into APIs from spreadsheet models, as well as Python Optalitix hosts the API, and SDK documentation is generated by the system for easy integration Experience monitoring is enabled using bespoke, real time dashboards Data is pushed into SQL databases that enable rapid model recalibration Model updates can be uploaded instantly and regression tested for robustness
- 8. Experience monitoring dashboards are customised in our insurer portals to track model accuracy 8 Checks on accuracy for propensity models Analysis of age experience Product recommendation accuracy measures
- 9. Our insurer portal screens provide data driven insight 9
- 10. What we do 10 We design systems for banks and insurers using machine learning We build innovative front-end portals for financial services Enable complex algorithms and integrations for financial clients Technology used and trusted by blue chip banks and insurers
- 11. Blue chip clients and partners 11