At Great Learning, I built a hybrid GPT-4o and Random Forest model that predicted learner attrition with 78% accuracy. I used SHAP to identify EMI payers as the highest-risk segment, supporting targeted dropout interventions.
I also built an LLM assistant that drafts replies to more than 300 support tickets per month, with low-confidence cases escalated for human review. For AlphaSwarm, I built an AI-powered algorithmic trading platform that turns plain-English ideas into deployable Python strategies and tests them in a realistic backtesting engine.

