In order to more efficiently allocate sales opportunities and prioritize resources, an AI solution was envisioned that would analyze relevant public information about a prospective lead along with historical win/loss rates from prior CRM opportunities in order to predict win rates for specific deals.
New leads captured in a CRM platform are sent into Contextual for AI processing. In a series of Contextual Flows, the company of the lead is extracted and public information from the website is analyzed and summarized based on specific prompts. All of this detail is stored as records in Contextual Object Types. The summarized details about the business are categorized using a classification LLM and then the entire data set is fed into a basic machine learning model trained on historic sales win/loss data and hosted on a third-party platform. The result is a win/loss prediction for the opportunity along with a model confidence score. All of that information is then passed back into the CRM platform to drive lead assignment and pipeline prediction.
