Jeaviit Bhaat

I build agentic AI systems in production: LLM pipelines, voice AI (TTS + STT), retrieval-augmented knowledge bases, prompt optimization loops, and evaluation frameworks. I don’t wait for tutorials; I ship things that work a month later.

As Founder & AI Engineer at GrowthOS, I designed and shipped production agentic AI voice agents on Retell AI—deploying inbound/outbound conversational pipelines with ElevenLabs TTS, real-time STT, tool use, and LLM-backed response logic for real clients in HVAC, medspa, and dental. I also built an automated deployment pipeline that scrapes a business website, constructs a RAG knowledge base, configures prompts, and deploys a shareable preview link in under 60 seconds.

I focus on reliability: I built evaluation frameworks for voice agents with scripted persona tests, root-cause analysis on failure modes, surgical prompt optimization, and regression checks before every redeployment. From there, I engineer production-quality full-stack apps in React/Vite/Tailwind/GSAP and experiment with Claude API + MCP tooling to improve structured tool calling and multi-step workflows.