Roark
The QA and observability platform for Voice AI that monitors live calls, runs simulations at scale, and turns failures into tests.
Roark
The QA and observability platform for Voice AI that monitors live calls, runs simulations at scale, and turns failures into tests.
Executive Summary
Roark is a seed-stage YC W25 company building QA and observability tooling for Voice AI agents — essentially "Datadog for Voice AI" — and it's entering the market at arguably the right moment, just as enterprise Voice AI deployments are scaling faster than testing infrastructure can keep up. The core thesis is sound, the team has directly relevant engineering pedigree, and the 10M minutes processed claim is plausible given named customers like Podium and Aircall, though it remains self-reported. The single biggest risk is structural: Roark operates one layer above the infrastructure platforms it integrates with (VAPI, Retell, LiveKit), and those platforms have every incentive and resource to build native QA features that would undercut Roark's standalone value proposition — a disintermediation threat that has already killed independent tooling vendors in adjacent markets.
Run your own diligence
Upload a pitch deck or paste any company URL to get a full AI-powered due diligence report in under 2 minutes.
Get started free →Free plan available · No credit card required