Hari santh

Built Pinecone vector databases and embedding-based RAG retrieval pipelines for 2 client AI assistants, improving domain-specific answer accuracy and reducing out-of-scope hallucinations. Delivered 3 production AI web applications integrating ML inference APIs with backend systems, cutting preprocessing-to-deployment turnaround by 40%.

Built supervised ML pipelines covering preprocessing, feature engineering, hyperparameter tuning, and 5-fold cross-validation for classification and regression tasks. Evaluated models using ROC-AUC, RMSE, and F1-Score across the full experimentation lifecycle.

Built reusable n8n automation modules connecting REST APIs with AI inference services for agentic workflow automation. Built AI-enhanced features for a VR/AR product management platform using React and Node.js, improving interaction efficiency by 15%. Built PDF log processing pipelines with automated analytics.

Grade: Second Class Upper Division

Pursuing a B.Sc. in Information Technology with a focus on Data Science, expected to graduate in September 2026.