The startup behind the viral AI coding app Cursor is snapping up prime expertise from AI enterprise startups in a bid to bolster its competitors with Microsoft’s GitHub Copilot and win over companies seeking to supercharge their workers with AI coding instruments.
In a single current case, Cursor maker Anysphere struck a deal to accumulate the AI-powered buyer relationship administration (CRM) startup Koala, two sources acquainted with the matter instructed TechCrunch.
As a part of the deal, Cursor will deliver on a number of of Koala’s prime engineers to construct out a devoted enterprise-readiness staff. Nonetheless, the whole Koala staff is not going to be becoming a member of Anysphere, and Cursor doesn’t plan to combine the startup’s core CRM product, a supply instructed TechCrunch.
Koala plans to close down in September, the corporate mentioned in a weblog publish printed on Friday. The announcement comes simply 5 months after Koala raised a $15 million Sequence A led by CRV, with participation from HubSpot Ventures, Recall Capital and Afore. Koala was practically 4 years previous, had roughly 30 workers in line with LinkedIn, and had labored with shoppers resembling Vercel, Statsig, and Retool.
Sources on this story requested anonymity to talk with TechCrunch about personal issues. Koala and its founders didn’t reply to TechCrunch’s request for remark. Cursor declined to remark.
The Koala deal paints an image of the 2 kinds of AI startups we’re seeing in 2025. There’s Cursor, a juggernaut of an AI software that’s rising so quick it’s beginning to encroach on the AI area’s largest gamers, together with Microsoft and Anthropic. On the identical time, there’s a rising variety of startups like Koala: B2B AI startups that appeared promising — with a co-founder from Meta and advisors like Jack Altman — however have shortly run out of steam.
Cursor is capitalizing on this disparity, leveraging middling AI startups as a way to construct out its personal enterprise choices. Anysphere additionally lately employed the CEO of cybersecurity startup Resourcely, Travis McPeak, to steer the corporate’s safety groups, in line with The Info.
Techcrunch occasion
San Francisco
|
October 27-29, 2025
These offers look lots like Massive Tech’s reverse-acquihires, resembling Meta’s current deal to rent Scale AI’s leaders. Very like in Meta and Scale’s deal, Cursor can now transfer shortly to construct out new enterprise segments whereas leaving questionable companies behind.
Cursor hopes that Koala and Resourcely’s expertise will assist it evolve from a private developer software that engineers quietly use at work and change into an enterprise‑broad platform that corporations pay massive contracts to entry. Most enterprises in the present day that provide workers an AI software select Microsoft’s GitHub Copilot, which works as an AI-powered extension to current built-in improvement environments (IDEs) resembling VS Code or JetBrains.
Cursor, which is a standalone AI-powered IDE, must beat out GitHub Copilot to strike offers in lots of instances, and it typically does in head-to-head assessments. Nonetheless, Microsoft has the higher hand within the enterprise because of its long-standing relationships with legacy corporations, in addition to its massive gross sales, safety, and assist groups.
Within the final 12 months, Cursor has decisively constructed out its go-to-market and gross sales staff — which now accommodates dozens of workers. A number of Cursor workers now spend their days visiting the places of work of Fortune 500 corporations and exhibiting them how they’ll combine Cursor’s AI instruments into their enterprise, in line with an individual acquainted with the matter.
Cursor’s enterprise push appears to be gaining traction. Anysphere mentioned it reached $500 million in ARR in June, and now works with greater than half of the Fortune 500, together with NVIDIA, Uber, and Adobe. A supply acquainted with the matter says that income has since grown, and an more and more massive share of that progress comes from enterprise offers.
However as Cursor competes with Microsoft, it additionally must fend off a rising area of threats.
Essentially the most urgent one is Anthropic, an important companion to Anysphere whose Claude Code product has grown quickly in current months. Cursor depends closely on Anthropic’s AI fashions to energy its personal coding merchandise. (As one among its largest clients, Cursor can also be important to Anthropic.)
On the identical time, Google simply scooped up the management staff of Windsurf, a significant competitor to Cursor within the AI-powered IDE area. Cognition, the maker of the AI coding agent Devin, acquired the remainder of Windsurf’s staff, probably providing a major elevate to each companies.
It’s value noting that every one of those are several types of AI instruments, however employers are inclined to view the merchandise equally: as AI instruments that may enhance the productiveness of software program engineers. These employers might need the suitable concept. In spite of everything, Anthropic, Microsoft, Cursor, and Cognition are all growing AI coding brokers which intention to automate workflows utterly, which can be the place the AI coding area converges.
You would possibly ask, why is everybody competing with everybody else to construct AI coding merchandise? Coding instruments have change into one of many first AI merchandise to search out “product market match” — the elusive purpose that makes the ears of enterprise capitalists perk up. AI coding merchandise are getting used day by day by hundreds of thousands of software program engineers, they usually have began to generate actual income.
Put one other means, the race isn’t nearly constructing the perfect AI coding software anymore. It’s about who can scale their enterprise operations the quickest whereas the market continues to be up for grabs. With the likes of Microsoft, Google, and Anthropic all shifting quick, Cursor’s acquisition technique might decide whether or not it joins their ranks or turns into one other startup that couldn’t scale quick sufficient.