Hi, I’m Gaurav Pooniwala.

Founder & AI engineer focused on LLM reliability.

Over the past decade, I've built production AI systems - from video generation used by 30M+ users (acquired by Adobe) to RAG and agentic workflows powering logistics and finance products.

I'm now focused on the challenges of LLM observability, human-in-the-loop review, and designing systems that stay reliable in high-stakes environments.

I also collaborate with teams as a consultant, advisor, or senior IC to help design and ship dependable LLM workflows.

How I work with teams

I partner with teams building or scaling LLM-powered products — especially where reliability, workflow design, and human oversight are critical.

Highlights

Built GenAI video systems used by 30M+ users (acquired by Adobe)

Designed RAG & agentic workflows for finance and logistics products

10+ years building production AI systems with real-world impact

Areas of Expertise

LLM Reliability & Observability

Failure modes, evaluation design, confidence scoring, HITL workflows.

RAG & Agentic Workflows

Retrieval, structured reasoning, routing, multi-step orchestration.

AI Workflow & Process Design

Balancing automation and human oversight across end-to-end pipelines.

Strategy, Leadership & Execution

Zero-to-one product thinking, cross-functional alignment, operational clarity.

Ready to build reliable AI systems?

Let's discuss how I can help your team move from prototype to production.