Sri Krishna Bharadwaj Eyunni

Built a plain-English analytics agent (Claude NLP-to-SQL) that lets restaurant owners query their own business data, with SELECT-only access across 9 tables, validation on every query, and per-query cost logging. Designed an AI-assisted ingestion pipeline that standardizes restaurant data from 6 file formats across 5 categories, validating every record and logging failures for data-quality review.

Deployed a full-stack web application on AWS using containerized FastAPI backend on ECS Fargate, static Next.js frontend on S3, and ECR-managed Docker images, with automated CI/CD via GitHub Actions using OIDC-based IAM authentication. Shipped a Graph RAG knowledge graph that cut LLM tokens by up to 84% (59% average), with Langfuse cost and quality tracking on every run.

Developed and validated a sequence-aware deep learning model (LRCN) for video-based human activity recognition, achieving 98.03% accuracy on temporal datasets. Published peer-reviewed findings at IEEE ICDABI 2024, validating model reproducibility and temporal classification methodology.

Pursuing a Master of Science in Data Science, Analytics and Engineering, expected to graduate in May 2026.

Completed a Bachelor's in Computer Science Engineering with a specialization in Data Science (AI and ML).