Find us at the conference entrance to see how the world's most accountable voice AI teams test, monitor, and prove their agents are deployment-ready.
Join our CEO for a panel discussion
Tuesday, Aug 4 · 3:00–3:45 PM · Bellini Ballroom 2102
CEO & Founder, Coval
Brooke Hopkins is the founder and CEO of Coval, the evaluation platform for voice and chat AI agents. Before starting Coval, she led evaluation infrastructure at Waymo. There she identified a striking parallel between self-driving systems and conversational AI: both succeed or fail on edge cases, long-tail behavior, and rigorous evaluation.
Today, Brooke is helping enterprises replace slow, manual QA with automated simulation and observability, so AI agents can be deployed and scaled with confidence.
Bellini Ballroom 2102 Tuesday, Aug 43:00–3:45 PM
Our CEO Brooke Hopkins joins data and engineering leaders from Ford, Cummins, 7-Eleven, and Devon Energy to dig into observability for complex AI deployments — tracking model performance in production, detecting drift and bias, tracing decision flows, and building the feedback loops that keep models honest at scale.
Agentic Live Demo Stage, Hall B Thursday, Aug 62:20–2:35 PM PDT
See how teams simulate thousands of real-world scenarios before launch, catch failures in production, and close the quality loop — all in one platform. Find us at the Agentic Live Demo Stage in Hall B.
Manual QA was built for a different era. Coval gives teams the infrastructure to ship voice AI they can actually trust.
Handles accents, interruptions, background noise, and tool calls. Everything that makes voice different.
Covers every stage of agent evaluation, from your first simulation to production monitoring at scale.
Human judgment is built into every eval. The more your team reviews, the sharper the whole system gets.
Trusted by teams putting voice AI in front of millions of customers.
Scale with Confidence
Simulation, monitoring, and human review work as one loop, so every production insight becomes the next test and each release improves agent performance.
Use voice AI simulation to pressure-test thousands of realistic callers, hard policies, noisy audio, and the workflows manual testing cannot cover.
Run production evals on live conversations and surface the patterns that show where resolution, safety, or experience is slipping.
Route the calls that matter to human QA, then turn reviewer judgment into sharper voice agent evals for the next release.
Voice AI evaluation across every lifecycle stage — from first simulation to production monitoring.
See it live→217%
Improvement in agent accuracy in 7 days
3.1M
Evaluation metrics run weekly
<15m
To first simulation via CLI
Schedule time with the Coval team at the conference for a live demo, a conversation about your use case, or just to connect.
