We started working on this in February, and incorporated shortly after our first YC interview in S13. We have about 5,000 lines of code written.
Our first product is an LTV optimizer for e-commerce: We are making a product to identify a business' most valuable customers at activation, and deliver higher conversion product recommendations per user, so as to catalyze a seller's rev-positive engagement cycle with their customers.
Sellers have figured out that having an online market is a huge revenue opportunity, and they even know that data analysis and recommendation systems directly improves their top-line revenue. The problem though, is that most businesses outside of the tech industry (and even tech companies) are not yet savvy enough to do their own data analysis--and when they do, it's at great cost, either by hiring and building internal capacity, or by spending large sums of money on data consultants.
From here, we believe there's a way to abstract machine learning and predictive analytics for all verticals. Just as Quid and Palantir have proven out high margin business models delivering people + software, we want to be the abstraction layer for companies that want to backtest their data to predict future behavior from their customers. Right now non-tech savvy companies are using descriptive statistics and punditry to make guesses about what to do next. We want to turn every company in to a data-driven organization. Think of it as the Palantir for the future.