Sushant Charaya

Built and launched four AI agent products (Validation Agent, Company Intelligence, Fraud Detection Agent, COI Agent) end-to-end from discovery and PRDs to engineering and client launch. For Uber, delivered the Validation Agent with 95% field-level accuracy, ~65% manual-review reduction, and 5× faster vendor onboarding; also achieved 91% fraud detection accuracy and automated certificate verificati

Scoped and built an LLM-based defect classification solution for Uber Eats Grocery using LangChain and RAG, reaching 89% accuracy and reducing manual audits by 80%+. Built Business Enrichment and conversational AI pipelines for merchant onboarding and sentiment analysis, cutting bug detection turnaround from 13 days to 1 day and reducing LLM token costs with FAISS + ChromaDB semantic search.

Led product analytics initiatives that boosted user engagement by 15% and customer satisfaction by 25%. Built ETL pipelines and Tableau dashboards for real-time product metrics and decision-making.

Analyzed business performance to improve margin from 7% to 10%, conversion rate from 82% to 89%, and Deal of the Day GMV contribution from 3% to 12%.

Earned a BS in Mathematics and Scientific Computing, graduating in December 2019.