Pratham Gangavalkar

I’m an AI/ML Engineer with hands-on experience across Data Science and Generative AI application development. I build end-to-end ML systems—from predictive analytics and classification to time-series forecasting and LLM-powered RAG—focused on accuracy, reliability, and scalable delivery.

In my role at TURYON Tech Labs Pvt. Ltd., I designed a hybrid code-understanding system (SyntaxLens) that combines rule-based AST parsing with LLM-driven semantic inference to explain Python code at configurable depth levels. I also built a FAISS-backed RAG retrieval layer that injects language-specific context at inference time, improving explanation quality without retraining the base model.

Previously, as a Data Science Intern, I worked through full ML project cycles, building preprocessing and feature engineering pipelines and translating model outputs into interactive dashboards for non-technical stakeholders. I gained practical experience with model versioning, performance monitoring, and deployment workflows across multiple deliverables.

I’m especially energized by NLP and GenAI systems that stay grounded—prompt engineering, embeddings, semantic search, and vector databases. I currently deploy RAG solutions using tools like LangChain, Gemini/OpenAI APIs, and Streamlit, and I’m actively seeking opportunities in AI/ML Engineering, Generative AI, and Data Science.