At Turing, I build and deploy LLM-powered apps, chatbots, and recommendation features for US startups. I designed LangGraph workflows with tool calling and evaluated 1,000+ AI agent trajectories for a frontier AI lab.
At Devsinc, I owned backend architecture for Mode Mobile's rewards app serving 1M+ users. I shipped OpenAI-powered content categorization and personalized messaging, and tuned database queries and schemas to cut slow endpoint response times by about 35%.
Earlier at Devsinc, I built core modules for SCMS, a multi-tenant certificate management SaaS, including renewal automation. At Sapling, I built request workflows, real-time chat, and notifications for Citizen Centric.
In my projects, I’ve built RegAssist, a multi-tenant agentic RAG platform with hybrid search, reranking, and source-cited answers, and AgentLens, an evaluation toolkit that checks agent traces and flags regressions. I also built AI assistants and agents for healthcare questions, code review, and natural-language analytics.
