Applied ML/Stable
Real-time UPI fraud detection over amount velocity, merchant diversity, geolocation entropy, and time-of-day features, with a moving threshold for a 0.01% base rate.
Feature engineering over transaction velocity and behavioral entropy, paired with a threshold-moving strategy so the classifier stays useful at a fraud rate of one in ten thousand.
Tell me what you're building, the constraints you're working with, and where it breaks. I reply within a day.