At CARNOT RESEARCH PRIVATE LIMITED, I engineered a WhatsApp chatbot using LLMs and RAG that served 83K+ users with 2K concurrent interactions. I also tuned backend APIs and async processing, increasing throughput by 35% while keeping latency under 300 ms.
At ARDEE FOUNDATION, I refined the Ask Atlas chatbot’s two-stage LLM pipeline and reworked conversation context handling to keep answers coherent across identified edge cases. At FUTURESTIC, I automated student onboarding and placement workflows with n8n pipelines and Claude-powered agents, cutting manual intervention by 25%.
I also designed ORQESTRA, a multi-agent system that analyzes third-party APIs and generates integration code, with a self-healing RAG validation layer. For PINGDAEMON, I built a monitoring backend for 40+ services with scheduled health checks and state-change email alerts.

