PranavSingh Patil

I've built production-grade AI systems across TheGrowthVector and Nikqik Tech, turning LLM and retrieval concepts into reliable backend products.

At TheGrowthVector, I engineered FastAPI and PostgreSQL AI backends, multi-agent workflows, and an NLP-to-SQL retrieval system using OpenAI APIs and PowerPlatforms. I also built distributed job orchestration with Celery, Redis, and Azure Service Bus, resolving queue-lock failures and improving long-running workload reliability.

At Nikqik Tech, I designed an end-to-end RAG pipeline for enterprise meeting intelligence, from transcript ingestion and parsing through retrieval and LLM inference. I built a custom .vvt transcript parser, MCP server and client plugins, and backend services using Python, JavaScript, MongoDB, and Azure.

My projects include Readify, MeshCards, and Debator-CLI—systems for document intelligence, AI-generated flashcards, and multi-agent debate orchestration. I enjoy combining strong systems design, cloud infrastructure, and practical AI workflows to build products with real-world impact.