Chamath Palihapitiya has stepped back into an operating seat for the first time since Facebook, taking the CEO role at 8090 Labs and announcing a $135 million Series A round led by Salesforce Ventures. The company is building what it calls Software Factory, an AI coding platform aimed squarely at the enterprise teams that cannot afford to “vibe code” their way into production.

The round closed June 29, 2026, with backing from a who’s who of tech circles: Jeffrey Katzenberg’s WndrCo, David Sacks’ Craft Ventures, David Friedberg’s The Production Board, and Jason Calacanis’s Launch. Angel investors include Palo Alto Networks CEO Nikesh Arora and Quora CEO Adam D’Angelo.

The Problem 8090 is Trying to Solve

AI coding assistants have taken off fast. But most of them were built for individual developers, not engineering teams inside banks, hospitals, aerospace companies, or government agencies. The result is a wave of “vibe coding” — AI-generated code that works well enough in a demo but cannot survive a compliance audit, a security review, or a production incident.

8090’s Software Factory is designed for a different category of customer. It brings people and AI agents into a single governed workspace where every decision, from intent to execution to deployment, is tracked, visible, and auditable.

That last word matters enormously in regulated industries. Healthcare teams face HIPAA. Financial services teams face SOC 2 and ISO requirements. Government contractors face FedRAMP. For these organisations, “the AI generated it” is not a satisfying answer when something goes wrong.

Software Factory promises to change that equation: not by removing AI from the loop, but by wrapping it in the controls that enterprise teams actually need.

Why This Investor Group Makes Sense

The investors are not random. Salesforce Ventures, the lead, has spent years watching enterprise software get disrupted from the outside. It has a front-row seat to what business software customers want next, and apparently it believes AI-native development tooling is going to take share from traditional enterprise software in a significant way.

The All-In connection (Chamath, Sacks, Friedberg, Calacanis) signals conviction from a group that is unusually aligned on where enterprise AI is headed. This is not a spray-and-pray round. It is a concentrated bet that the “production-grade AI coding” category is real and that 8090 can own it.

The angel roster reinforces the enterprise thesis. Nikesh Arora runs Palo Alto Networks, a company that thinks deeply about how code vulnerabilities propagate through enterprise infrastructure. He is not investing in a toy.

What This Signals for Enterprise Software

The timing matters. Cursor, the AI coding assistant, was just acquired by SpaceX for $60 billion. GitHub Copilot has 1.8 million paid subscribers. Microsoft has integrated AI coding into its entire enterprise stack. The market for AI-assisted software development is no longer speculative.

What 8090 is doing is carving out the regulated-industry segment of that market and saying: those customers need something different. They need governance, auditability, and enterprise-grade oversight baked in from the start. A $135 million Series A led by the CRM category leader suggests the market agrees.

What This Means for Business Leaders

If you run a team that builds or relies on internal software, this trend matters to you in three ways.

The gap between “it works in a demo” and “it runs in production” is widening. AI can generate code faster than ever. But deploying that code safely, especially in regulated environments, requires exactly what 8090 is building: structured oversight, audit trails, and human-in-the-loop governance. The tools that win enterprise deals will be the ones that close this gap.

“AI coding” is not a single category. There is consumer-grade AI coding (Cursor, GitHub Copilot’s free tier), developer productivity AI (Copilot Pro, Replit Agent), and enterprise-grade AI software development (where 8090 is playing). Conflating these when making technology decisions will lead to the wrong tool for the job.

The build-vs-partner question is getting more complex. As AI coding tools mature, some businesses will be tempted to build everything in-house with AI assistance. But building and maintaining AI-powered software at enterprise scale requires expertise that most internal teams do not yet have. The right answer for most organisations is still a combination: a strategic partner for the complex custom work, and the right tooling for the routine stuff.

For Enterprise DNA clients, our Omni Apps service sits at that strategic layer: identifying where custom AI applications will create real competitive advantage, building them properly, and ensuring they are maintainable over time. The 8090 story validates the market. The implementation is still where the hard work happens.