I build AI/ML systems and agentic pipelines, which mostly means I spend my days convincing software to make decisions on its own and then panicking when it actually does.

• 🧠 Currently training multi-agent systems to be smarter than my code reviews
• 🐛 Bug-to-feature conversion rate: surprisingly high
• ☕ Runs on caffeine, vibes, and the occasional working pip install
• 🤖 My agents have better uptime than my sleep schedule
• 📍 Tamil Nadu, India — usually somewhere between Colab and a stack trace

"If it works, don't touch it. If it doesn't, blame the LLM."

⭐ If you read this far, you deserve a star too — feel free to drop one on a repo.