ZoomInfo Agent Teams now available, no new contract required

On September 30, ZoomInfo announced that it had acquired DoubleO.ai, an agentic orchestration platform founded in 2024 and built to help teams create reliable, multi-agent automation.

Agent Teams is built on DoubleO.ai's orchestration architecture, which ZoomInfo announced it had acquired the day before. It is available now in GTM Studio, with no new SKU or contract required. For the sales operations or RevOps leader at a company already on ZoomInfo, that could move the first question away from whether to buy multi-agent automation and toward whether these agents run a sales play the same way twice. DoubleO.ai CEO Derek Osgood said the company was built to run complex workflows reliably, every time.

Agent Teams ships inside GTM Studio with no new SKU or contract, so a ZoomInfo customer can test DoubleO.ai's reliability claim on its own accounts right away.

A sales play that survives a two-week gap

ZoomInfo's launch announcement on Business Wire describes Agent Teams as a "real-time evaluation architecture." In plain terms, each agent has its own role, instructions and tools. Agents pass along the context from every earlier step, and each one picks its next action based on the current state of the account record, not a fixed script.

The champion example shows why that matters. Picture a former champion who takes the CRO job at a target account. Agent Teams first confirms whether that company is already a customer and whether it matches the seller's ideal customer profile. Next, it pieces together what the champion did in the original deal, identifies who sits on the new buying committee, and writes outreach that draws on the existing relationship.

For operators whose champion tracking stops at a CRM task, the pricing-page restart could be the behavior to test first.

The second example begins once a deal is closed-won. In both plays, the agents share context from every prior step.

Why a data company bought an orchestration engine

citybiz described the deal as extending ZoomInfo's strategy, from supplying GTM data and intelligence toward using that information as context for AI systems that carry out broader revenue workflows. Dealroom put the bar more bluntly. As agents move into daily workflows, "reliable enough to run" is becoming the test that decides which vendors win.

ZoomInfo's acquisition announcement also points to GTM.AI, which it calls its headless context layer. Citybiz and Investing.com both reported that financial terms weren't disclosed.

Putting the 20% versus 10x claim to work

The strongest objection to this story is that reliability is still described mostly by the companies selling it. Caroline Burnett, sales operations leader at Silverfort, said in the launch announcement that Agent Teams let her scale RevOps and improve both the quality and the results of her GTM work. That's an encouraging data point, and it's a single one. With no new SKU or contract required, trying Agent Teams could be easy, but measuring the results would likely fall to buyers themselves.

What the agents read, and who signs off

Agent Teams coordinates agents across a company's systems, so security and data governance teams belong in the evaluation, not just sales ops. For companies with sensitive deals or regulated customers, the useful question is which records the agents can read, and whether particular accounts can be excluded.

The champion play deserves the same scrutiny. It drafts outreach based on a past relationship, so the setting to check is whether agents send messages or queue them for a rep to approve. The launch materials describe drafted outreach. How each customer sets that control will shape how quickly the play moves from RevOps testing to every rep's daily work.

ZoomInfo's own launch names the gap it wants to close: useful agent projects that stay in one person's workspace and never reach the front line. The sign that Agent Teams is working will be more named customers like Silverfort describing plays that run across every rep's accounts, not just inside a RevOps sandbox.