8090 Labs Raises $135M to Automate Software Development with AI Agents

8090 Labs, an enterprise AI coding startup co-founded by Chamath Palihapitiya and CTO Sina Sojoodi in January 2024, has closed a $135 million Series A led by Salesforce Ventures. Palihapitiya, who had been on the board rather than operating day-to-day, is stepping into the CEO role as part of the round. The company's "Software Factory" platform aims to let teams of human engineers and AI agents build and maintain enterprise software together.

Last updated August 4, 2026: verified the funding-announcement quote against primary reporting, added 8090's March 2026 EY.ai distribution partnership, and added pricing and positioning context against Cursor, GitHub Copilot, and Cognition's Devin, none of which were in the original report.

What Happened: 8090 Labs' $135M Series A

The Software Factory platform covers more than code generation: it handles requirements refinement, architecture capture, and work planning, feeding AI agents structured project context rather than one-off prompts, before generating code that human engineers review and integrate. The company's initial focus is regulated verticals like healthcare and finance, where custom workflows and compliance requirements make off-the-shelf software a poor fit.

Beyond Salesforce Ventures, the round drew from all four hosts of the All-In podcast: Palihapitiya himself, Jason Calacanis (LAUNCH), David Sacks (Craft Ventures), and David Friedberg (The Production Board), along with Jeffrey Katzenberg's WndrCo and angel investors Nikesh Arora (CEO of Palo Alto Networks) and Adam D'Angelo (CEO of Quora). Salesforce's strategic lead signals a plausible integration path into its existing enterprise platform footprint.

Announcing his return to a full-time operating role, Palihapitiya said: "AI can write code. The hard part of enterprise software is keeping fifty agents and a hundred engineers changing the same complex system every week without it pulling apart." Separately, he described the funding round as the moment he'd been waiting for since leaving Facebook to return to full-time operating work.

Where Software Factory Fits Against Cursor, Copilot, and Devin

8090 is entering a market with an established leader: GitHub Copilot had roughly 4.7 million paid subscribers as of early 2026, by most trackers somewhere in the high-30s to low-40s percent of the paid AI-coding-tool market. Cursor, the developer favorite for fast, AI-native editing, prices its Teams plan at $40 per user per month. Cognition's Devin, following its acquisition of Windsurf, prices Devin Teams around $80 per month plus $40 per full seat.

Software Factory's pricing, reported around $200 per user, is aggressive by comparison, and the company isn't competing on IDE features. Its differentiation is scope: Copilot and Cursor generate and complete code inside an existing workflow, while 8090 positions itself across the full development lifecycle, from requirements through validation, which is a smaller, more enterprise-specific category also occupied by Devin and newer entrants like Tembo. A March 2026 partnership with EY.ai gives 8090 concrete enterprise distribution, putting the platform in front of tens of thousands of EY consultants rather than relying solely on direct sales.

The structured-context approach Software Factory is built around addresses a specific, well-documented limitation of code-completion tools like Copilot and Cursor: they generate code well within the scope of what's visible in the current file or a limited context window, but they don't inherently track how a change ripples across a large system's broader architecture, business requirements, or compliance obligations. In a fifty-agent, hundred-engineer environment, the kind of scale Palihapitiya described in his launch quote, that gap compounds quickly: an AI agent that generates locally correct code without understanding a system's full requirements can still introduce integration bugs, break compliance constraints, or duplicate work another agent already completed elsewhere in the same codebase. Feeding agents structured requirements and architecture context up front, rather than relying on each prompt to carry that context implicitly, is 8090's specific answer to that coordination problem.

Why It Matters

Software development remains a bottleneck for enterprise digital transformation, and the $200/user price point only works if 8090 can prove the full-lifecycle approach saves more engineering time than a $40/month code-completion tool. The EY.ai distribution deal is the more concrete signal to watch than the funding total: enterprise software adoption usually depends more on who's selling and integrating it than on how much capital a vendor has raised.

The healthcare and finance focus is a deliberate wedge, not an incidental starting point. Both sectors carry regulatory requirements, HIPAA in healthcare, a range of financial-services compliance regimes depending on jurisdiction, that make audit trails, requirements traceability, and change documentation a hard requirement rather than a nice-to-have. A code-completion tool that generates a function correctly but leaves no record of which business requirement it satisfies or why a particular approach was chosen creates compliance risk in those industries, which is exactly the gap 8090's requirements-and-architecture-first workflow is positioned to close, and why the company is starting there instead of competing head-on with Copilot in less regulated, more price-sensitive markets first.

What Happens Next

8090 will use the funding to expand Software Factory's capabilities and grow past its current healthcare and finance customer base. The EY.ai partnership gives an early read on real enterprise adoption; whether AI-generated code at $200/user actually outperforms cheaper code-completion tools on complex, regulated systems is the open question the next year of customer results will answer.

Final Takeaway

8090's bet is that enterprise buyers will pay a premium for full-lifecycle AI development over cheaper code-completion tools, backed by real distribution through EY.ai rather than funding alone. Whether that premium holds up against Copilot's incumbency and Devin's lower price point is the real test ahead, not the size of this round.

Key Points

AI-Native Development Platforms

8090 Labs represents a category some are calling "AI-native" rather than AI-augmented development: traditional IDEs add AI features like code completion to existing workflows, while AI-native platforms rebuild the process around AI as the primary code author and humans as reviewers and integrators. That shift is contested among software engineers, some see it as removing tedious work, others worry it reduces programming to prompt engineering and oversight in ways that can degrade architecture quality over time if review discipline slips.

FAQs