Venkata Kaushik Mitta

AI Engineer with 2.5+ years of hands-on experience building production-oriented Generative AI, RAG, and NLP systems using Python, LangChain/LlamaIndex-style orchestration, OpenAI APIs, and Hugging Face transformer workflows. Skilled in designing end-to-end RAG pipelines — document ingestion, chunking, embedding generation, vector indexing (FAISS, Chroma), semantic retrieval, and LLM response grounding — and in measuring system quality through custom evaluation scripts and before/after benchmarking. Experienced in deploying AI services via FastAPI/REST APIs, Docker, and cloud infrastructure (AWS, Azure, GCP). Track record of measurable impact: 30% improvement in retrieval relevance (achieved by optimizing chunking, embeddings, and retrieval logic, validated through custom evaluation benchmarks), 40% faster model release cycles, and 15% gains in prediction accuracy.