Pit

Pit is a Stockholm-based enterprise-AI-agents startup founded by the founders of e-scooter company Voi, with a $16 million seed round led by Andreessen Horowitz in May 2026.
Pit

Pit is a Stockholm-based enterprise-AI startup that builds AI agents for back-office and service functions in industrial verticals, founded in early 2026 by Adam Jafer, Fredrik Hjelm, and Filip Lindvall, the co-founders of European e-scooter company Voi. The company emerged from stealth in May 2026 with a $16 million seed round led by Andreessen Horowitz with participation from Lakestar, Nordic family offices, and angels including senior executives from US technology companies. The product premise is to package agentic AI as a deliverable internal-tooling capability for European industrial enterprises, with a vendor-agnostic stack and an emphasis on enterprise governance, certification, and auditability.

At a glance

  • Founded: Early 2026; pilot testing began mid-January 2026, public launch in May 2026.
  • Status: Private.
  • Funding: $16 million seed round (May 2026), led by Andreessen Horowitz (partners Alex Rampell and Gabriel Vasquez) with Lakestar, Nordic family offices, and senior US technology-executive angels participating.
  • CEO: Adam Jafer.
  • Other notable leadership: Fredrik Hjelm (co-founder, currently dual-hatting as Voi CEO), Filip Lindvall (founding engineer), Andreas Hjelm (engineer).
  • Open weights: None. Pit is a deployment-and-orchestration company, not a frontier-model trainer.
  • Flagship products: Pit Studio (a no-code authoring environment for enterprise employees to guide AI through internal processes) and Pit Cloud (a deployment surface for AI-generated software with enterprise governance, certification, and audit-log requirements).

Origins

The founding team came out of Voi, the European e-scooter and shared-micromobility company that the same three founders started in Stockholm in 2018. Voi had grown to approximately 1,000 employees operating in 13 countries by the time Adam Jafer departed in early 2026, with the company reportedly reaching profitability in 2024 and producing a "strong 2025" in the operator's framing. Voi has been mentioned in European-tech-industry coverage as a potential 2026 or 2027 IPO candidate.

The Pit founding decision was driven, in the founder's framing, by an observation that the generation of AI models reaching production in late 2025 and early 2026 had crossed a capability threshold past which agentic systems could do "most of what junior engineers used to do." Adam Jafer made that argument publicly on LinkedIn before founding Pit and later qualified it to acknowledge that scaled implementations still require a meaningful human mix. The thesis underlying the company is that the structural opportunity is in the enterprise back-office automation layer (the work that internal IT, operations, customer-service, and shared-services teams do) rather than in customer-facing chat assistants or general-purpose foundation-model wrappers.

Fredrik Hjelm remains chief executive of Voi and is involved in Pit as a co-founder rather than as a full-time operator; the founder's framing is that Hjelm will be "less hands-on for the time being." Filip Lindvall, the third Voi co-founder, joined Pit as a founding engineer. Andreas Hjelm, Fredrik's brother, is part of the engineering team. The founding cohort also includes engineers recruited from iZettle (the Swedish payments company acquired by PayPal) and Klarna, two of the most prominent post-2010 European technology companies and the dominant source of senior engineering talent in the Stockholm ecosystem.

Mission and strategy

Pit's stated mission is to deliver agentic AI as a capability that enterprise procurement organisations can buy rather than as a tool that individual developers integrate. The strategic premise is that the AI category around 2026 has two distinct customer-side problems: the model-quality problem (largely solved by the closed-weight frontier labs and the strong open-weights challengers) and the productionisation problem (how to deploy agentic systems with enterprise governance, audit logs, certification, and integration with the customer's existing IT stack). Pit addresses the second problem rather than competing on the first.

The vendor-agnostic stance is structurally important to the strategy. Pit's product positioning explicitly avoids tying customers to a single AI model or cloud provider, and the operator-level framing emphasises that the company "uses different AI and cloud vendors depending on clients' preferences." The framing has two strategic uses: it sidesteps the procurement-side concern about single-vendor lock-in that has been a recurring objection to closed-weight AI procurement, and it positions Pit as a long-lived integration layer that survives any individual frontier model's commercial trajectory.

The target-vertical framing is heavy industrial Europe. Adam Jafer has stated publicly that the company is "going after industrials, and there's plenty of that in Europe." Initial pilot customers are reported in telecom, healthcare, and logistics. The geographic framing leans on a secondary thesis: that European industrial procurement is increasingly demanding "EU models running on EU compute" for sovereign and regulatory reasons, and that Pit's vendor-agnostic posture is the right shape for that procurement pattern.

Models and products

Pit has two primary product surfaces.

  • Pit Studio is a no-code authoring environment for enterprise employees (operations leads, customer-service managers, shared-services directors) to guide AI agents through internal processes. The framing is that the people who understand the process best are the operators themselves, and the authoring surface should let them encode that knowledge into agent behaviour without writing code or sitting through a long professional-services engagement.
  • Pit Cloud is the deployment-and-runtime surface for AI-generated software. It provides enterprise governance (RBAC, audit logs, change-management workflows), industry certification (Pit has indicated SOC 2 and ISO 27001 are on the roadmap), and runtime monitoring. The framing is that AI-generated software has the same integration and compliance requirements as human-authored enterprise software, and that the deployment-and-runtime layer needs to satisfy those requirements without imposing a development workflow that defeats the speed advantage of agentic generation.

The product scope is deliberately bounded to back-office and service functions. Adam Jafer described the scope as "nothing customer facing, no conversational AI, just pure back-office, service, and support functions." The decision to avoid customer-facing surfaces is a strategic differentiation against the broader Voice AI and customer-service AI categories, which are crowded with both well-funded incumbents (Intercom, Zendesk, Sierra) and frontier-lab consumer products.

Benchmarks and standing

Pit has not published model-benchmark numbers as of release. The company's product positioning is on integration depth, governance, and time-to-deployment rather than on model-quality leadership, so the standard frontier-model benchmarks (Artificial Analysis Intelligence Index, LMArena, GPQA, SWE-bench, HumanEval) are not the appropriate measurement instrument. Independent benchmarking for the company's actual contribution layer (enterprise-deployment time, certification coverage, customer-side adoption metrics) will likely emerge over 2026 to 2027 as the company adds customers and shares case-study data.

Leadership

  • Adam Jafer (Chief Executive Officer): Voi co-founder, departed Voi in early 2026 after a seven-year tenure during which he ran engineering. The CEO role at Pit is his first chief-executive position.
  • Fredrik Hjelm (Co-founder): Voi co-founder and current Voi chief executive. Joined Pit as a co-founder while remaining at Voi in a less-hands-on capacity for the new venture.
  • Filip Lindvall (Founding Engineer): Third Voi co-founder. Joined Pit at the founding-engineer level, with primary responsibility for the engineering organisation alongside Adam Jafer.
  • Andreas Hjelm (Engineer): Brother of Fredrik Hjelm; part of the founding engineering cohort.

The senior-engineering hiring profile is concentrated in former iZettle and Klarna staff, which is the standard Stockholm-ecosystem pattern for new technology companies. Specific names beyond the founding cohort have not been publicly disclosed. Active hiring at launch focused on solution engineers and forward-deployed engineers, the customer-side delivery roles that enterprise software companies use to ship and integrate at the customer site.

Funding and backers

The $16 million seed round of May 2026 was led by Andreessen Horowitz with partners Alex Rampell (founder-investor on multiple enterprise-software exits) and Gabriel Vasquez (a16z's growth-stage partner with focus on European deals). Lakestar, the European venture firm with strong Stockholm-ecosystem ties, participated. The remainder of the round was filled by Nordic family offices and US-technology-executive angels, none of whom were named publicly at the announcement.

The fundraising framing is interesting on two dimensions. First, the founders' framing was that they "did not need the money to get going, but wanted the strongest backers we could find," which positions the seed round as a strategic-relationship round rather than a runway round. Second, the speed of the fundraise (the founders stated they did not spend much time with other firms) reflects the talent-leads-capital pattern documented in the diaspora map at the lower-frontier tier: the Voi-founder credential plus a clean enterprise-AI thesis compressed the time-to-term-sheet in the way that frontier-lab tenure compresses it for senior AI researchers leaving Anthropic, OpenAI, or DeepMind.

The a16z relationship dates from the Voi years; Alex Rampell and the Voi founders had stayed in touch in the intervening period, which is the structural condition that produced the term-sheet velocity.

Industry position

Pit's industry position is at the intersection of three current categories.

First, enterprise-AI deployment platforms (the category occupied by companies like Sierra, Glean, and the customer-AI offerings from Salesforce, ServiceNow, and Microsoft). Pit's differentiation against this category is the back-office-and-service vertical focus and the vendor-agnostic stack, both of which the broader enterprise-AI category has not consistently emphasised.

Second, European-sovereign-AI infrastructure (the category that includes Mistral's enterprise offerings, the broader French AI-sovereignty agenda, and emerging European-cloud alternatives). Pit's positioning leans on the "EU models on EU compute" framing but does not commit to a single sovereign-stack vendor, which is consistent with its vendor-agnostic strategy.

Third, the Stockholm AI-scene cohort. Lovable, a Stockholm-based AI-engineering startup that emerged in late 2024 and 2025, is the closest local peer; the two companies are positioned at different layers of the stack (Lovable on developer-tools, Pit on enterprise-deployment) but share the Stockholm-ecosystem talent base and the a16z-as-European-partner posture. Industry coverage has characterised both companies as part of a broader Stockholm cohort that the venture market is increasingly treating as a serious European tech hub.

Competitive landscape

  • Sierra (US, conversational AI): Bret Taylor's enterprise-AI agents company. Customer-facing rather than back-office, and US-centric in customer base. Different end-state target market.
  • Glean (US, enterprise search and AI): Enterprise-search-and-AI deployment company with broad US enterprise customer base. Overlaps on the AI-deployment-platform category but not on the back-office-process-automation focus.
  • Salesforce Agentforce (US, customer-AI platform): The incumbent customer-AI offering. Strong distribution into existing Salesforce customers but customer-facing rather than back-office.
  • Microsoft Copilot Studio (US, enterprise-AI authoring): The deepest-pocketed competitor on the no-code authoring surface. Microsoft's distribution advantage is large but the European-sovereign-stack requirements that Pit emphasises are not Microsoft's natural positioning.
  • Mistral AI enterprise offerings (France): The French enterprise-AI offering with sovereign-stack credentials. Different layer of the stack (Mistral is a frontier-model vendor; Pit is an integration-and-deployment vendor) but the two compete for the European-enterprise-procurement budget at the highest level.
  • Voi-ecosystem competitors: Bird, Lime, and other shared-micromobility companies the Pit founders previously competed against in the e-scooter market. Not relevant to the AI category per se, but the operator-side experience from that competitive market shapes the founders' execution priors.

Outlook

Open questions and watchable signals over the next 6 to 18 months:

  • First named enterprise customers. The pilot-customer roster in telecom, healthcare, and logistics has not been publicly disclosed. The first round of named customer case studies will be the central signal on whether Pit's integration-depth thesis converts to actual production deployments at meaningful contract values.
  • Adam Jafer's full operator transition. The CEO transition from Voi-engineering-leader to Pit-chief-executive is the founder's first sustained CEO tenure. How the operator transition plays out (specifically, whether the Voi-operator experience translates into the AI-deployment-platform category, where customer dynamics and procurement cycles differ) will shape the company's 2026 to 2027 trajectory.
  • Series A timing and round size. The Series A round, if and when it closes in 2026 or 2027, will reveal whether the pilot-customer base produced enough revenue or named-customer traction to justify the kind of valuation that the European AI ecosystem has been pricing recent enterprise rounds at.
  • Vendor-agnostic positioning under scale. Pit's "we use different AI and cloud vendors depending on the customer" framing is structurally easier to deliver at low volume than at high volume, where integration depth with any single vendor creates leverage. Whether the company maintains the vendor-agnostic posture or whether it converges toward a primary model-and-cloud partner as it scales will be visible in the integration-architecture-side messaging through 2027.
  • Fredrik Hjelm's eventual transition. Hjelm's dual-hat role (Voi CEO plus Pit co-founder) is a transitional arrangement. The eventual transition (back to Pit full-time, or to a fuller Voi commitment) will signal which company the founder team treats as the long-term primary vehicle.

Sources

About the author
Nextomoro

Nextomoro

nextomoro tracks progress for AI research labs, models, and what's next.

AI Research Lab Intelligence

nextomoro tracks progress for AI research labs, models, and what's next.

AI Research Lab Intelligence

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to AI Research Lab Intelligence.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.