Sai Kiran

I’m an AI/ML Engineer specializing in LLM systems, RAG architectures, NLP, and computer vision. I focus on building production-ready ML pipelines that deliver reliability, measurable quality, and real-world usability.

In a Google Partner Project internship, I designed and deployed an automated ML pipeline integrating REST APIs (Meta/Facebook Ads → CRM) with real-time ingestion, deduplication, and lead scoring. I reduced manual processing overhead by 45% through preprocessing and feature engineering workflows, and improved throughput by optimizing SQL queries and backend indexing.

Across my RAG work, I’ve led prompt engineering and embedding optimization to drive performance—maintaining 99% uptime for a production RAG system handling 500+ daily concurrent queries. On a legal-domain RAG project, I achieved strict grounding with <5% hallucination rate, improved clause-level relevance by 28%, and built citation generation that increased lawyer trust and adoption by 3x.

I also build supporting systems for evaluation and deployment, including MLflow-based monitoring and FastAPI inference with auto-scaling (sustained <1.2s response time under 50 concurrent users). I’m motivated by end-to-end ownership—turning model capabilities into dependable products.