The Social Capital founder is going full-time CEO at his own portfolio company, a move that signals just how serious the race for enterprise software automation has become.
For years, Chamath Palihapitiya was one of the most vocal voices in venture capital, the kind of investor who shaped conversations from a podcast seat rather than an operator's chair. That changed yesterday. With 8090 Labs closing a $135 million Series A led by Salesforce Ventures, Palihapitiya announced he's stepping down from the board and into the CEO role full time, running the AI coding startup he founded in January 2024.
The round pulled in some familiar names: David Sacks' Craft Ventures, Jason Calacanis' LAUNCH, Jeffrey Katzenberg's WndrCo, and David Friedberg's The Production Board all participated. So did Palo Alto Networks CEO Nikesh Arora and Quora CEO Adam D'Angelo as angels. That's not a random collection of investors. That's the All-In podcast inner circle, plus enterprise-credentialed operators writing checks. Salesforce Ventures leading the round is the telling detail: Salesforce sells to the Fortune 500, and it doesn't write $135 million Series A checks for consumer tools.
8090 Labs builds a product called Software Factory, an AI coding agent aimed squarely at corporate development teams rather than individual developers. The pitch is straightforward: enterprises don't need an AI tool that writes clever code in a sandbox. They need one that ships production-ready software with the audit trails, compliance controls, and governance infrastructure that heavily regulated industries demand. Healthcare, financial services, aerospace, energy, and the US government are 8090's target customers, according to the company. Ernst and Young has already partnered with 8090 to launch what the firms are calling EY.ai PDLC, a product delivery lifecycle platform built on top of Software Factory.
Here's what the AI coding market actually looks like right now. GitHub Copilot, backed by Microsoft, has over 4.7 million paid subscribers and solves around 56% of SWE-bench benchmark tasks. Cursor, the VS Code fork that hit a reported $50 billion valuation, sits just behind on the benchmarks. Both are excellent tools for developers who want to move faster on their own machines. Neither was built for the compliance officer sitting three floors above the engineering team.
That gap is where 8090 is positioning. Individual developer productivity and enterprise software delivery are genuinely different problems. A developer using Cursor to ship a side project doesn't need an immutable audit log. A bank rewriting its loan origination system does. The question isn't whether AI can generate the code. It's whether the organization can trust the process that generated it enough to put it in production. 8090's bet is that most enterprise buyers will pay serious money for that answer, and the $135 million raise suggests at least a cohort of experienced investors agrees.
Palihapitiya's decision to take the CEO seat himself is worth reading carefully. Venture capitalists don't typically give up their GP roles to run portfolio companies unless they believe the opportunity is large enough to justify the opportunity cost, or unless they think the company needs something only they can provide. Possibly both apply here. As TechCrunch reported, Palihapitiya had been serving as a board member since founding the company; moving to full-time CEO eighteen months in is a meaningful escalation of personal commitment.
The company says it will use the funding to expand its technical team, scale hiring broadly, and invest in high-performance compute and infrastructure. That last part matters. Running AI coding agents at enterprise scale, where a single software delivery pipeline might involve thousands of files, strict access controls, and zero tolerance for hallucinated dependencies, is computationally expensive in a way that individual developer tools are not. You can't do it cheaply.
What 8090 is ultimately building is a claim about who controls software production in large organizations over the next decade. If AI coding agents get good enough, the question stops being which IDE a developer prefers and starts being which platform a CTO signs a multi-year contract with. That's a much larger deal size, a much longer sales cycle, and a much more defensible position once you're embedded. Salesforce built a $30 billion business on exactly that dynamic. It's not surprising they want a seat at the table for the next version of it.
Palihapitiya has spent years arguing on the All-In podcast that incumbents are underestimating how completely AI will restructure enterprise software. Running 8090 is the bet behind the argument, and $135 million is now on the line to prove it.
Also read: Companies spending most on AI are hiring faster, not cutting headcount, a billion-job study finds • Meta secretly sent fake teen accounts to flood rivals' chatbots with thousands of crisis prompts • South Korea is treating nuclear construction speed as an AI infrastructure problem and it's right
