Ask my CV anything - out loud.
This is a demo voice agent that answers questions about Jamie Curnow, built in a weekend as a portfolio piece. Press call, ask it about my work, and hear it answer in real time.
It’ll tell you up front that it’s an AI. Access is limited to invited domains.
Production voice stack
Twilio Media Streams into a real-time pipeline - streaming speech-to-text, LLM, and text-to-speech. The same architectural shape as a production voice agent, scoped small.
Evals & observability
Every turn is measured: latency percentiles, confidence, grounding, token cost. The thinking is in the instrumentation, not just the demo.
Responsible automation
Up-front AI disclosure, answers grounded strictly in the CV, and confidence-based escalation to a human when it shouldn’t guess.