Khushi Verma

I’m a Python & AI/ML engineer with 2+ years of experience building GenAI applications, RAG pipelines, and Python backends (Django/Flask). I’ve delivered enterprise AI solutions using LangChain, LlamaIndex, and LLM orchestration—cutting manual effort by up to 80% and improving operational efficiency at scale.

At KPIT Technologies Limited, I architected a multi-agent LLM pipeline with Python, LangChain, LLM APIs, and RAG-based vector retrieval—reducing hallucinations and improving response accuracy by 40%. I also built Python + FastAPI AI tooling to generate DoD/DoR/acceptance criteria and tasks from raw requirements, adopted by 5+ teams and reducing manual documentation effort by 70%.

I enjoy turning complex requirements into scalable, reliable systems. I’ve implemented MLOps with Vertex AI for model deployment, versioning, monitoring, and continuous evaluation, reaching 99.5% uptime, and developed FastAPI microservices handling 1000+ concurrent requests with <500ms latency.