Company Performance Metrics
Joe Lacy: FounderPast Role: WagePit, Founder
The problem we’re solving: Supervised ML alone doesn’t tell the full story. Unsupervised ML finds new patterns in your data and is better at anomaly detection. Our tool finds brand new patterns in data which is important when dealing with changing macroeconomic conditions. This is especially necessary when dealing with fraud, predictive modeling,
and rule optimization.
Why we’re the ones building it: We’re making something that our data scientists wish they had when they worked at their prior companies (Capital One, Tesla, Fidelity, DARPA/DoD). Because we regularly faced the problems we are solving for, we’re the best primed to build this tool for users like us at companies all over the world.
So, what can you expect? 1. We’ll alert you within 24 hours when there has been a fundamental change in the makeup of the data coming into your models.
2. Our anomaly detection is best-in-class. We R&D-ed and integrated our own method of ensembling multiple unsupervised ML models together to improve anomaly identification by 35% over single model anomaly detection.
3. Our tool finds patterns in your data without needing that data to be fully (or perfectly) labeled, which is needed for supervised ML.
