AI/ML Engineer
NITEC
Mar 2025 — Present● Improved the accuracy of the RAG-based Egov AI chatbot by 5% through a reranking system ● Advanced RAG pipeline for an agricultural subsidies chatbot, implementing caching to reduce token usage by 25% ● Developed benchmarks to evaluate RAG pipelines, LLMs, and embedding models ● Deployed LLMs on CPU/GPU infrastructure using Docker and vector database servers; created APIs for a text-to-SQL bot