Coval raises $28M to build the testing layer for voice AI

Read the original article →

What happened

Coval raised a $28 million Series A led by Norwest, with Base10, Twilio Ventures, and Y Combinator participating, bringing total funding to $31 million. The company builds simulation, observability, labeling, and review infrastructure for voice and chat agents.

Founder Brooke Hopkins brings simulation experience from Waymo, applying autonomous-vehicle testing discipline to voice AI. Customers include Zoom, Deepgram, and more than 60 organizations, which have run tens of millions of evaluations. Coval cites more than $7 billion invested in voice AI in the first quarter of 2026.

Why it matters

Voice agents fail in ways text agents do not, from interruptions to accents to timing. Systematic simulation and evaluation are how those systems move from convincing demos to reliable production.

The funding is part of a broader wave backing evaluation infrastructure, a sign that reliability, not raw capability, is the current bottleneck.

MintedBrain take

Teams shipping voice agents underinvest in testing at their peril. Borrowing simulation practices from higher-stakes domains is sound, and evaluation coverage should scale with how much you trust the agent to act unattended.

References

This article was originally published at Tech Startups. For the full piece, read the original article.

Discussion

  • Loading…

← Back to News