Krinish Radadiya — Engineering Manager

I am an AI Engineer and Engineering Manager who builds LLM‑powered systems end‑to‑end – from data and retrieval pipelines to production‑ready APIs and evaluation frameworks. With 5+ years in software engineering and 2.5+ years focused on AI development, I combine deep hands‑on work with practical product thinking and team leadership. Started my AI journey in classical machine learning, working with NumPy and Pandas on datasets with millions of records. Built a strong foundation in data quality, feature engineering, and performance at scale. One early solution automated a client's manual analysis workflow, saving 30+ hours/week and improving decision consistency. Shifted to Generative AI, designing RAG systems with advanced chunking strategies, embedding pipelines, and ranking models. Worked extensively with Milvus, Faiss, Chroma, S3 Vector, pgvector/Postgres – tuning retrieval for low latency and production reliability. Optimized one RAG project to cut latency by ~30% and token usage by ~20% through better chunking and model selection. Work across OpenAI, Google, Anthropic, Mistral providers and built a provider orchestrator that intelligently routes queries based on intent, cost, and performance. Created voice/calling agents with TTS/STT models, focusing on context switching and LLM orchestration. Comfortable with Postgres/SQL, MongoDB for scalable backends. Specialise in deployment + measurement: model serving with vLLM, production inference, and chatbot evaluation frameworks benchmarking relevancy, latency, and completeness. Current stack: LangChain, FastAPI, Langfuse, N8N and more.. I deliver quality code, clear communication, open-source contributions, and love tackling new technical challenges while mentoring teams to ship robust, observable AI features.

Stackforce AI infers this person is a SaaS-focused AI Engineer with strong leadership in developing scalable AI solutions.

Location: Vadodara, Gujarat, India

Experience: 4 yrs 11 mos