We're hiring, so if you're impressed by these notes, check out our careers page: https://www.infer.com/careers.html or email me directly jh<lastname>@infer.com
We are working on a similar concept.
We aim to predict - Customer behaviors based on a function of their Geography, previous CRM notes, tags, liking and tastes. We call that as Interest-Graph and store them as nodes. We identify patterns.
We don't focus on leads; as we don't have much data. Rather repeat buyers, these are previous customers. We try to predict buying-behaviors from data in structured, unstructured sources, papers, CRM's and ERP.
So far we are able to predict: what campaigns might work on a group of customers, based on our graph data.