Koala AI Built the Best Intent Signal Platform in B2B Sales, Then Walked Away

The Macro: Sales Teams Are Drowning in Signals They Can’t Read

Every B2B sales team has the same problem. There are too many leads, not enough context, and the reps who close deals are the ones who somehow figure out which accounts are actually ready to buy right now. The rest of the team is guessing. That guessing costs real money.

The intent data market has been growing because of this exact pain. Tools like Bombora, 6sense, and ZoomInfo have built enormous businesses around the idea that you can observe buyer behavior and predict purchase timing. Bombora aggregates content consumption signals across a co-op of publishers. 6sense uses AI to map anonymous buying activity to accounts. ZoomInfo layers contact data on top of firmographic profiles. They all work, to a degree. But they all share a common limitation: they sit outside your product.

The signals that matter most for a SaaS company aren’t which whitepapers a prospect downloaded from a third-party site. They’re what that prospect did on your website. Which docs page they read. Which pricing page they visited three times in one afternoon. Whether they just signed up for a free trial and immediately started poking at the API. Those first-party signals are gold, and most sales teams had no clean way to access them.

This is the gap Koala walked into. And for a few years, they owned it.

The Micro: Thirty Signal Sources, One Prioritized List

Koala came out of Y Combinator’s Winter 2022 batch with a thesis that was almost aggressively simple: unify first-party and third-party intent signals into one view, score accounts automatically, and hand sales reps a prioritized list of who to call today. Not next week. Today.

The founding team was led by Tido Carriero as CEO, alongside co-founders Matt and Netto. Carriero had previously been VP of Engineering at Segment (which sold to Twilio for $3.2 billion), so he understood the data plumbing problem intimately. The team knew that most intent data tools failed not because the signals were bad but because the signals were scattered across too many systems for any rep to synthesize in real time.

Koala pulled from 30+ signal sources and 30+ enrichment data sources. It tracked website visits, documentation page views, feature usage patterns, and CRM activity. It layered third-party intent data on top. Then it ran AI scoring to surface the accounts showing buying behavior right now, not the ones who might be interested someday.

The product had a few features that made it sticky. Koala Plays let teams set up automated workflows triggered by specific signal combinations. If a VP of Engineering from a target account visited the pricing page twice and then opened a support doc about enterprise SSO, that would trigger a play that alerted the right rep with the right context. Koala Coach used AI to suggest next steps for each account based on the signals it was seeing.

The customer list was strong. Vercel, Retool, Hightouch, Sanity, Dolby, Deepgram, Clay, Postman, and others. These are developer-focused companies where the buying signal lives in product usage, not in a Gartner download. Koala claimed it generated over $1 billion in pipeline for its customers collectively, and at a $15 million Series A, the economics looked right.

The product was SOC2 Type II certified, which matters when you’re asking companies to pipe their customer behavior data through your platform. Integration points included native connections to CDPs, data warehouses, product analytics platforms, and the usual CRM suspects.

I spoke with a few people who used Koala at mid-stage startups, and the consistent feedback was that it was the first intent tool that felt like it was built by people who understood how modern SaaS companies actually sell. It wasn’t trying to be a data warehouse. It was trying to be the thing a rep looks at before they pick up the phone.

The Verdict

Here’s where the story takes a turn. In mid-2025, Koala announced it had been acquired by Cursor, the AI-powered code editor that has been tearing through the developer tools market. The entire founding team joined Cursor. Koala’s product was given a shutdown date.

Carriero wrote something honest in the announcement that stuck with me: “The joy we felt for how we were building began to overshadow the joy we felt for what we were building.” That’s a founder saying the technology they were using to build Koala (Cursor, specifically) became more interesting to them than the problem Koala was solving. It’s rare to see that kind of candor.

The product is technically still accessible as of this writing, but the clock is running. Customers have been migrating. The team is at Cursor now, working on enterprise readiness alongside another recent acquisition, Resourcely.

So why write about a product that’s shutting down?

Because what Koala built was genuinely good, and the category it defined is not going away. First-party intent signals are still the most valuable data in B2B sales. The companies that were using Koala now need to find alternatives. Some will go back to Bombora or 6sense. Some will try to build it internally. Some will wait for the next startup that tries to do what Koala did.

The lesson here is that a great product and real traction don’t guarantee survival if the founding team finds a different mission they care about more. That’s not a failure. It’s just a specific kind of ending.

If you were a Koala customer, I’d be looking at Clearbit (now part of HubSpot), Warmly, or building a lightweight version with your existing CDP and a bit of custom scoring logic. None of them are Koala. But the signals are still there, waiting for someone to read them.