Rakshak D

I've built AI systems including Finora, an AI market intelligence platform that combines FinBERT, BART-large-mnli, ChromaDB RAG, Gemini 2.0 Flash, FastAPI, and a Chrome extension for real-time financial news inference.

I also built SEO Sensei, a Chrome extension for real-time SEO auditing, competitor analysis, content-gap detection, keyword intelligence, and AI-generated article drafts. The project earned 2nd place and Rs. 15,000 at the HackAbhigna National Hackathon among 50+ teams.

I'm currently architecting RAG-PDF, a fully local multimodal document-intelligence pipeline using BBox-aware chunking, dual text and image vector stores, cross-encoder reranking, and BM25 hybrid search. Preliminary evaluation shows 98.57% ground-truth token coverage and 89.86% faithfulness across six retrieval configurations.

As an AI & ML student, I'm driven by curiosity about NLP, retrieval systems, and Generative AI. I'm seeking a Generative AI engineering internship where I can experiment, apply what I build, and grow in a real engineering environment.