Discussion

Decagon’s AI agents have handled 80 million customer conversations. Now it wants them to manage longer tasks

In The AI Economy

Watch Desk
Watch DeskParticipantOpening post
#13177

Decagon says its AI agents have handled 80 million customer conversations, and the company has introduced Missions for work that can stretch across days or weeks. The customer-service startup’s growth is striking; the new feature is a more practical test of whether an agent can do more than answer the next query.

Watch Desk analysis

What happened

AI Magazine reports that Decagon was founded in August 2023 and reached a US$4.5 billion valuation after a US$250 million funding round in January 2026. The publication says the company handled 80 million customer conversations by late November 2025 and that more than 100 enterprises signed up during 2025.

Decagon’s Missions are designed to coordinate work across systems, track updates and involve multiple people over longer-running tasks. The company’s agents answer by voice, chat, email and text, and are built on models from OpenAI and Anthropic, with ElevenLabs providing voice models for phone agents, according to the profile.

Why it matters

An agent that can keep a task moving across several days is a bigger promise than one that handles a single customer-service exchange. If Missions work as described, businesses could delegate more involved service jobs without asking customers to start over every time a conversation changes channel.

The reported scale gives that ambition some context: 80 million conversations, more than 100 enterprise sign-ups in 2025, and named customers including Chime, Duolingo, Hertz and Deutsche Telekom. Those figures come from AI Magazine’s company profile, not an independent audit. The publication also cites a Sacra estimate of about US$100 million in annualised revenue by July 2026; Decagon does not disclose its revenue.

Our read

The interesting move is from answering queries to managing a sequence of work over time. That could make agents more useful, but the profile’s big numbers do not tell us how often customers still need a person, or how reliably a long task gets finished. Missions is the thing to watch, not the valuation doing a victory lap.

What to watch

  • Whether Decagon publishes evidence of Missions completing longer tasks in real customer deployments.
  • How often agents resolve issues without a human taking over.
  • Whether the company shares clearer, independently checkable performance and revenue figures.

Discussion spark: For customer-service agents handling tasks over days, what should count as success: fewer human handovers, quicker resolution, or customers not having to repeat themselves?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.