Voiceflow shipped upgrades to its observability suite to make it easier than ever to understand agent behavior, including transcript keyword search and knowledge base logs joining playbook, workflow, and tool logs in Log Visibility settings.

Voiceflow is the AI agent platform for CX owners. Companies deploying AI customer experiences today face a bad set of options: hand the whole thing to a managed-service vendor, wait for a legacy suite to modernize, or drop it on engineering and hope nothing breaks — three different ways of renting your customer experience from somebody else. We think your CX is a proprietary asset. You should own the interactions you deliver, the outcomes you drive, and the technology behind both. Voiceflow gives CX and product teams one place to build, test, deploy, and improve AI agents across voice and chat. Your engineers get headless APIs and MCP for backend routing and architecture, while your CX team designs in an intuitive build interface with complete control. With a robust observability suite, either side can trace exactly what happened when an agent does something unexpected. What that looks like in practice: • AI agents that orchestrate both generative reasoning playbooks and deterministic workflows, so refunds and billing changes never get improvised • Connections to the CCaaS, ticketing, CRM, telephony, and model providers you already run, with no migration required • Pricing based on compute instead of "resolutions," so your costs track real usage rather than a vendor's margin CX teams at Turo, GetYourGuide, StubHub, Superloop, and Trilogy run high-volume support on Voiceflow, hitting 80%+ automation rates across more than 100 billion tokens processed. Renters stay tied to someone else's roadmap, pricing, and results. Owners compound every workflow they build, every skill their team picks up, and every model they swap in. The future belongs to the owners. Book a demo at voiceflow.com