Pit’s $16M bet: AI-built software for enterprise workflow chaos

The workflow problem it wants to replace

Pit is aimed at the messy operational layer inside companies: processes that are often coordinated through spreadsheets, email inboxes, manual handoffs and disconnected SaaS tools ^[4][6][14].

Coverage of the company cites examples such as campaign management, logistics coordination, approvals and inventory processes ^[6]. These are the kinds of workflows where rules, exceptions, approvals and data movement can vary heavily from one organization to another, making them difficult to fit neatly into a generic software product ^[6][15].

How the “AI product team” model is supposed to work

Pit’s model reverses the usual enterprise software pattern. Rather than asking employees to adapt to a rigid product, Pit says it builds operational systems around existing workflows, approvals and data flows ^[14][15].

In practice, the company’s approach can be understood in four steps:

1. Learn the operation. Pit says it studies how a company works before building the system that supports it ^[14].
2. Build tailored internal software. Its platform is positioned as a way for enterprises to create custom, production-grade software for internal operations ^[1][4].
3. Automate manual coordination. The company is targeting workflows now handled through spreadsheets, inboxes, manual processes and fragmented SaaS tools ^[4][6].
4. Move beyond copilots. Pit’s stated ambition is not just to provide AI chat assistance, but to deploy operational software around workflows, approvals and data flows ^[15].

One report describes two product components: Pit Studio for building company-specific systems and Pit Cloud for secure, compliant infrastructure ^[12]. That detail matters because Pit is not simply pitching AI recommendations; it is trying to make AI-built internal systems usable inside enterprise operations ^[12][15].

Where the $16 million fits

Pit announced its public launch alongside a $16 million seed round led by Andreessen Horowitz ^[4][14]. Reported participants include Lakestar, the Stena and Lundin families, Pit’s founders and executives from companies including OpenAI, Anthropic, Google, Deel and Revolut ^[1][4][14].

The clearest reported use of the funding is continued platform development and global expansion ^[11]. In other words, the round gives Pit capital to test whether its “AI product team” concept can become a repeatable enterprise software business rather than a one-off custom development service ^[11][14].

Who founded Pit

Pit was founded in 2025 by Adam Jafer, Filip Lindvall, Fredrik Hjelm, Anton Öberg and Fredrik Olovsson, according to startup coverage ^[3][16]. The company’s launch materials and reporting describe the team as connected to Voi and Klarna, with some coverage also citing iZettle experience ^[3][14][16].

That operator background is relevant to the company’s pitch. Pit is not only selling AI generation; it is selling the idea that enterprise software should understand how work, approvals and data actually move across teams ^[14][15].

The real test: replacing workflows, not just tools

Pit’s launch is best read as an ambitious thesis, not proof that AI-generated software has already replaced enterprise systems at scale. The available sources document the company’s funding, public launch, positioning and target use cases; they do not independently prove broad production adoption across large enterprises ^[1][4][14].

The practical hurdles are significant. Enterprise buyers will still need confidence in integration, governance, security, reliability, user adoption and long-term maintenance. Pit’s bet is that an AI-native platform can absorb enough of that complexity to make custom internal software faster and more practical than stitching work together across spreadsheets, inboxes and rigid SaaS products ^[4][11][15].

If the model works, Pit could point to a broader shift in enterprise software: from companies adapting their operations to generic tools, toward tools generated around the way each company already works. For now, its $16 million seed round makes it an early, well-funded test of that idea ^[1][14].