AI Engineer

Published in Genpact, 2025

Building and shipping GenAI and agentic AI/ML solutions to streamline complex workflows and drive operational efficiency.

  • Improved an LLM-based price claims-reconciliation pipeline (Azure OpenAI, LangGraph) through prompt engineering, chain orchestration and iterative validation, improving the claim pass rate by 30%.
  • Rebuilt product matching with TF-IDF, fuzzy and cross-encoder semantic matching, reducing LLM-driven matching by 65%.
  • Built a retail pricing tool: CatBoost sales forecasting by price, size and attributes; scenario simulation and grid-search optimization; and a natural-language-to-SQL analytics layer (Gemini, LangGraph).