Eagle Eye Analytics is an analytics company offering advanced predictive modeling to the Property and Casualty insurance market. The company was formed 4 years ago and has been marketing a machine learning SaaS product for the last three years. The company prides itself on selling solutions to clients the way they want to buy them so they do everything from helping with additional analytic capacity to supplying a packaged analytic solution. Two key things drive the company:
- Analytics in insurance are significantly behind those in other industries
In particular they target the weaknesses of generalized linear models, the main tool used in insurance, and are very focused on delivering fine-grained segmentation models (these are often very simple and so easy to describe e.g. in business rules). Typical results have found improved lift of 2X or more over traditional methods.
- Strategically, EagleEye chose a SaaS delivery model in order to make powerful analytics accessible by any and all P&C carrier, not just the large players. Companies simply upload their data, run modeling tools or have Eagle Eye run them and get results.
Eagle Eye Analytics performs a lot of data cleansing and organization in the cloud once data is uploaded, so the data set often becomes the best set of data that the organization has. This means that clients do data mining on this data but they also use web-based tools to slice and dice it for “softer” analysis. The data mining tools themselves are aimed at a more analyst level (rather than a PhD mathematician) so there is a fair degree of automation and cross-checking built in. Their approach revolves primarily around a view of profitability (Loss Ratio improvement) but is also useful in other areas. Different analytic methods are available, some common and some specific to different areas. Within each product (Profit Talon, Rate Talon and Growth Talon) there are different business goals that have implications for the analysis of data.
Part of the Eagle Eye delivery approach is to focus on how to operationalize their models. The system is quick, with lots of pre-built understanding of insurance data and its typical problems, so models get produced in 30-60 days. The models are typically very easy to encode (their secret sauce is really to generate good segmentation) so many customers take the segmentation and implement it in their own business rules management system. Organizational change is still an issue though so Eagle Eye works through the implementation and monitoring process to make sure clients achieve the business results they were looking for.
Eagle Eye Analytics is investing in R&D around the feasibility and benefits of pooling data from clients to start building industry-wide models but today they are focused on a client’s own data. The most sophisticated carriers are looking for new techniques, new ways to get value out of their data so they use Eagle Eye Analytics as an add-on to their existing models. Mid-sized and smaller companies are working with Eagle Eye in place of other analytic services to answer operational questions and perform sophisticated analytics.