Harris Warren

About

I’m an AI Product & Program Leader - Scientist & AI Operator - who ships Voice AI, LLM systems, and robotics from lab to live. I cut latency, land customer-loved features, and align complex, multi-org programs to executive decisions. Recent work includes shipping a studio-recorded voice model that now powers 3M+ connections and winning 3 AI hackathons (1st Place, Meta XR; Grand Prize, Google DeepMind Gemini 3). Amazon Devices and Services Led a multi-org GenAI portfolio across voice AI, information services, safety, and robotics with 36 weekly stakeholders across the US, UK, Poland, and India. Shipped a studio-recorded voice to production on the AWS speech platform with more than 95 percent naturalness and less than 3 percent critical pronunciation error. Reduced robot voice latency by 120 ms and cut end-to-end voice application latency from 500 ms+ to 266 ms average with p95 at 289 ms. Built agentic analytics on Bedrock and LLMs that compressed triage from days to seconds across 15+ teams. Regular L8 briefings and an L10 presentation for a program-critical decision. Contract extended from 6 to 9 months prior to company-wide reductions. Battery Movement Solo full stack AI developer

Experience

Education