GOAL: Client was web scraping over 10,000 companies / mo into their funnel that broadly fit their investment criteria. However, they lacked a predictive model that could identify best targets for the team to spend their time pursuing for potential investment.

​

INSIGHT + ACTION: Developed a predictive ML-powered model that learned from and trained on 2 investment types: A) good outcomes (theirs & others), and B) bad outcomes. Back tested the model against historical investments and identified 95% of companies they should not invest in, and score-ranked the remaining 5%. Analysts now use the model to rapidly score and prioritize over 10,000 new top of funnel companies per month and confidently focus on the top 500 scored opportunities.

​

BUSINESS OUTCOME: Machine learning used to increase sourcing efficiency by over 20x. Reduced time to term sheet submission by over 25% .