Hi everyone! I'm Deependra Rajput, a Full Stack Developer and AI Engineer from Gwalior, India, with 3+ years of experience shipping products end to end.

What I work on

I'm a Software Development Engineer at Synthlane, where I build:

  • LLM-powered applications: prompt design, conversational workflows and AI features inside real products
  • AI voice automation: inbound and outbound calling agents that handle thousands of conversations a day
  • RAG systems: retrieval over product docs, knowledge bases and FAQs
  • Scalable backends: Python, FastAPI, Node.js, PostgreSQL and Redis, with React and Next.js on the front end

What I'll write about here

I want to share the practical side of production AI, the stuff that only shows up once real users arrive:

  • Cutting LLM latency (streaming, caching, prompt slimming)
  • How an AI voice agent is actually put together
  • FastAPI patterns that hold up under load
  • RAG that answers the question instead of just retrieving text

Say hi

If you're building with LLMs, voice AI or FastAPI, I'd love to connect. Drop a comment with what you're working on, or ask me anything about shipping AI to production.

Thanks for reading! 🚀