At Maincrafts Technology, I co-designed autonomous AI agents that automated multi-step workflow pipelines by 30%. Under senior guidance, I also implemented LoRA fine-tuning on domain-specific LLMs, reducing hallucination rates by 25% and improving contextual accuracy.
At Codec Technologies, I developed an end-to-end RAG pipeline with LangChain, Hugging Face, and FAISS. I optimized batch document ingestion to cut vector indexing latency by 20%.
On my projects, I built an enterprise document QA platform that achieved 92% answer relevance, and a self-hosted MLOps pipeline with drift detection and automated retraining. As AI/ML Lead at Google Developer Student Clubs (GDSC) — CGEC, I organized hands-on workshops and mentored junior engineers through machine learning pipelines.
