I care most about one thing: making AI radically cheaper to run. I turn hard problems into clean, reliable systems — lately, systems that deliver the same AI outcomes at a fraction of the cost.
💸 Reducing the cost of AI — cutting token usage through intelligent, efficient context management, and moving enterprise & AI-startup workloads onto smaller open-weight LLMs. Same quality, dramatically lower cost.
- 💡 Deep in low-cost AI: context engineering to squeeze more out of every token, and open-weight / small language models as a serious alternative to frontier-model bills.
- 🤖 Building toward agentic systems that stay cheap at scale — where smart context and right-sized models matter more than raw model size.
- 🌱 Currently exploring Quantum Computing and pushing on Small Language Models (SLMs).
- 💬 Ask me about cutting AI cost, context engineering, open-weight LLMs, agentic development, and large-scale system design.
- 🏢 By day, I lead Platform & Intelligence engineering at ONDC — building agentic intelligence for commerce at India scale.
- 📍 Based in NCR / Gurgaon, India · on GitHub since 2013.
Always happy to trade notes on cost-efficient AI, context engineering, and open-weight models. ✨