Decagon Adds AI Phone Support

Voice turns Decagon from a chat automation vendor into a full customer contact layer. That matters because the biggest, most expensive support queues still start on the phone, where companies pay humans or BPOs to verify accounts, handle returns, and resolve billing issues. By pairing Decagon’s action layer and procedures engine with ElevenLabs’ real time speech stack, Decagon can keep one workflow running across chat, email, and calls instead of forcing support teams to manage separate systems.

• The hard part is not just sounding human. It is completing real support work safely. Decagon’s voice agents plug into call center infrastructure and use the same Agent Operating Procedures that power chat and email, so the phone agent can authenticate a user, look up an order, process a return, or handle a charge dispute inside the same governed workflow.
  1 sacra 2 elevenlabs
• This pushes Decagon closer to Sierra and beyond the Intercom model. Sierra built early momentum around autonomous chat and voice, while Intercom bundles AI into a broader help desk with 99 cent per AI resolved ticket pricing. Decagon’s move into voice signals that AI support leaders are converging on owning the full resolution, not just deflecting text tickets.
  3 sacra 4 sacra
• The ElevenLabs partnership shows where value is splitting in the stack. ElevenLabs provides the speech quality, latency, and voice infrastructure, while Decagon owns the enterprise logic, integrations, and outcomes. That is the same pattern showing up across voice AI, where specialist audio vendors supply the ears and mouth, and application companies own the workflow and customer relationship.
  1 sacra 5 sacra 6 sacra

The next step is that phone support stops being a separate channel and becomes another entry point into the same AI agent. As voice quality improves and enterprises get comfortable letting agents take backend actions, Decagon can expand from inbound support into outbound collections, onboarding, and retention calls, which materially increases both its product surface area and its revenue base.