Avatar for Kalepa

AI-powered decision support for commercial underwriters

Jobs at Kalepa

We are working on a number of challenges across data engineering, data science, product engineering, and infra. We are building a platform to ingest and normalize, at scale, a vast number of very distinct structured and unstructured data sources. These range from web data to government data to private company details. One of the core technical challenges is building a business identity graph to map data sets to each other in a performant way that scales to millions of businesses — with thousands of relevant data points per business. On the data science from, the core challenge is to identify the signals that are predictive of risk across a number of potential perils (e.g., fire, theft, lawsuits, injury). This can range from applying NLP to text sources (e.g., reviews) and computer vision algorithms to satellite images, we are working with cutting-edge open source models and also building and training our own. Finally, we need to build beautiful, intuitive software for the insurance underwriters who we are looking to empower. These professionals are used to difficult to use, slow, and low quality software. We will change that.