We use opensource frameworks (h2o, xgboost, lightgbm, catboost, pytorch, tensorflow) for our solutions. We create flexible modules for different problems. We develop both difficult models for the best predictions and light models with the minimum number of predictors for the quickest count.
Analysts and managers often associate AI with the black box. We refute this statement. Our product allows looking into the depths of AI insurance models.
No more need to spend hundreds of hours developing web services. Our solutions – working services with implemented models – could be deployed in a few simple steps on the company's infrastructure or in the cloud.
Years of work. Dozens implemented models.
Now we are ready to present our new product that incorporates experience and knowledge on solving business problems by analytical methods.
For more than 10 years we have been developing models for insurance companies. Our achievements are improved loss ratios for hundreds of segments, increased sales in several companies, automated dozens of business processes.
You can find Insolver python library on github: https://github.com/MindSetLib/insolver
We can solve the tasks of pricing, customers’ scoring, and LTV estimation, managing outflows, anti-fraud, and increasing cross-sales. Innovative tasks like fast loss adjustment, car damage recognition by photo, and optimization of the call-center have become typical with us.
And all these types of models became accessible with our product — Insolver.
We take some concepts as a basis in a product:
What we are solving: decrease time-to-market for insurance models from days and months to minutes, transparency of each calculation for each quote, claim, and policy, sharing ideas inside the company.
Our secure environment allows us to work in the powerful cloud http://mset.space and do research and deploy models in minutes. Any analyst will become a data engineer, actuary, data scientist, and ml engineer in one click with our product.
Tens of different research reports and dashboards support the analytical decision process. With our solution, we provide unique educational content for analysts each week.
And all of these things will become accessible for analysts with Insolver.
You can find Insolver python library on github:
Responsible for BI, analytics, optimization problems, economics and education program. Strong experience in GLM models built on open source technologies.
Responsible for team development,
NLP and CV models in insurance, project management, financing and legal aspects of the product.
Responsible for graphical models, classical ML, various types of boosting. Strong experience in math, data science and actuarial calculations.
Responsible for product software and technological development. Makes our system work 24/7 in any condition.
Insolver library changes pricing for an insurance company. In 2018 we as a insover team decided to create a product that helps insurance companies to solve certain calculation and deployment problems. As analysts and participants Читать дальше…
How a middle company can increase sales using Insolver platform
Frank Shikhaliev
+ 7 926 379 01 23