I architect AI systems that bridge cutting-edge LLM reasoning with high-throughput production infrastructure.

I'm an AI Engineer specialized in Generative AI, Agentic Workflows, and Multi-Modal Systems. At Adda Education, I've designed and scaled production AI pipelines that automate core operations: architecting ScriptIQ to automate over 1,000+ educational scripts saving ₹60+ Lakhs annually, building Gradi.AI for multimodal video quality evaluation across 10,000+ instructional sessions, and delivering broadcast automations for 200+ channels that reduced production times from 30 minutes to 2–5 seconds (a 6x throughput leap).

My engineering philosophy centers on cost-efficiency, low-latency, and reliability. When designing parallel agentic audit architectures, I scaled automated evaluations across 326,000+ audio calls under strict API rate limits, achieving ~85% accuracy while slashing per-call costs by 58% and tripling speed. From sub-300ms real-time conversational voice agents to large-scale DynamoDB-to-BigQuery migrations reducing infrastructure costs by ~95%, I build end-to-end AI software that stands up under real-world production load.