Kovva's Agents Plug Into 50 Media Buying Tools To Automate the Gaps Between Platforms

Agentic AI startup Kovva has built agents that plug into roughly 50 tools media buyers already use — DSPs, search and social platforms, verification partners, Slack and Teams — with no new dashboard to learn. The startup's COO and co-founder, Tanja Mimica, laid out the approach on this week's episode of AdExchanger Talks.

Her core argument: every major ad platform is building AI to optimize itself, but that solves the wrong problem. Media buyers spend their days hopping between systems — pulling reports from a DSP, a social platform and an ad server, reconciling mismatched numbers, QA-ing campaigns and troubleshooting line items that fail to deliver.

"Meta is building AI and getting really good at optimizing Meta, and The Trade Desk is really good at optimizing The Trade Desk," Mimica said. But the "majority of the work of a media buyer lives between the platforms."

And that in-between work, she says, is exactly what nobody is automating.

How does a buyer start working with an agent?

Kovva's on-ramp is deliberately low-friction. Buyers begin with a daily or weekly briefing sent to email, Slack or Teams that recaps the past 24 hours across their accounts.

From there, they cross-check the briefing against what they see in the platforms, ask the agent follow-up questions and, once the agent proves itself, schedule recurring tasks and hand off more complex workflows. Mimica puts the timeline for that shift at a minimum of three to four weeks.

Trust builds gradually. Buyers remain cautious about delegating anything high-stakes or hard to undo — setting up a campaign with a large budget, for example.

"Which I think is perfectly understandable," Mimica said. "As someone who's trafficked campaigns before, I wouldn't either."

Who controls the guardrails?

Kovva builds its safety rules outside its agents. People set the rules and limits, and an agent does not have the authority "to raise its own budget limit or approve its own change," Mimica said.

Every "write" action — the term for an agent changing something inside a platform, as opposed to just reading data — requires human approval.

"We're planning on keeping that in place for a long time," Mimica said.

In short: for now, agents at Kovva can read freely but cannot act without a human signing off.

What is Mimica's path into ad tech?

Her route to ad tech was unconventional. Mimica studied animal physiology in Australia and worked in veterinary practice management before landing in the industry. The episode covers that journey, including how she used a baby name book to name her first ad tech startup.

She also addresses the human side of adoption: the AI holdouts and internal champions who can make or break rollout inside an organization, and how to pitch the business case for agentic AI beyond the raw number of hours saved.

Mimica will be honored at AdExchanger's 2026 Top Women in Media & Ad Tech Awards, with the gala set for November 4 in New York City.

The takeaway in plain words: Kovva is betting that the next layer of ad tech automation is not inside any single platform but in the messy connective tissue between them — and it is keeping humans in charge of every change an agent wants to make.