whoami

I'm Jyotiraditya Singh — most people call me Jyotir. I build AI systems and backend infrastructure, mostly in Python and Go.

My GitHub bio says "ADHD programmer; semi-prime example of jack of all trades, master of none," which is only half a joke. The through-line in what I build is systems that keep working when a piece of them stops: cross-vendor failover in Loom, replayable steps in OrderMesh, a voice agent in DAWA that still completes the call when the model dies.

experience

AI Developer Intern at Prometrix AI (Aug 2025 – July 2026). I authored Loom there, and it was adopted company-wide as the standard LLM/TTS/image integration layer across 11 projects. I also built AIVA, a generative-engine-optimization platform that tracks how often brands surface in AI-assistant answers to real buyer queries, and shipped an AI Podcast Studio — an LLM + TTS pipeline that generates multi-speaker podcasts with streaming output and personas assigned dynamically across a cast.

Smaller and more satisfying: benchmarking cosine, dot-product and L2 against Manhattan distance for a domain-specific retriever, then adopting the winner for +9% Recall@5. And cutting a dashboard from 1s to 400ms with Redis caching and query tuning.

education

B.Tech in Computer Science Engineering, University of Petroleum and Energy Studies (UPES), Dehradun — Aug 2022 to May 2026.

things I've Achieved :)

• Smart India Hackathon 2023 — National Finalist
• WebCrux 2024 Web3 Hackathon — Winner
• CSI Hackathon — Runner-up (solo)
• Sarvam Buildin' hours — Top 100 out of 6.2k

what I actually know

Languages — Python (primary), Go, Java, JavaScript, C++, SQL, Bash

AI & LLM systems — RAG, AI agents, voice agents (STT/TTS), vector databases, retrieval eval, OpenAI/Anthropic/Gemini APIs

Backend & APIs — FastAPI, REST, WebSockets, PostgreSQL, Redis, RabbitMQ, MongoDB, webhooks

Infrastructure — Docker, Kubernetes, Linux/UNIX, CI/CD, Nginx, Git