The AI Coworker Arrives: How Two Former Meta Engineers Built a $15 Million Business in 10 Weeks by Putting an Assistant Inside Slack

WARSAW — May 20, 2026 — When Fryderyk Wiatrowski and Peter Albert met inside Meta's AI division, they had no intention of starting a company. They were engineers, not entrepreneurs. Their evenings were spent tinkering — building small AI agents that could handle the repetitive parts of knowledge work that nobody enjoys. Writing reports. Pulling data from scattered systems. Setting up dashboards. The kind of tasks that fill calendars and drain energy without producing anything anyone would call meaningful.

What emerged from those late-night experiments was not supposed to be a business. It was a tool the two built for themselves — an AI agent that lived inside Slack, understood the company's tools and workflows, and could be messaged like any other colleague. "Viktor was supposed to be a small experiment," Wiatrowski wrote this week. "It became the AI coworker 10x'ing real businesses."

The experiment just landed $75 million. On May 19, Viktor announced a Series A funding round led by London-based Accel, with participation from Bek Ventures, Kaya VC, Inovo VC, and Tenacity Capital. The round — the largest ever raised by a Polish-founded company — valued Viktor at several hundred million dollars and attracted an extraordinary roster of angel investors, including Slack co-founders Stewart Butterfield and Cal Henderson, Synthesia CEO Victor Riparbelli, Vercel founder Guillermo Rauch, Instacart co-founder Max Mullen, and executives from Google DeepMind, Figma, and ElevenLabs.

The numbers that drew them in are difficult to dismiss. Three months after its public launch in February 2026, Viktor reached a $15 million annualized revenue run rate — a milestone that took Slack itself nearly three years to achieve. More than 2,000 organizations now use the product across sectors including e-commerce, technology, and professional services. The company says Viktor connects to over 3,000 workplace tools, from Google Drive and Notion to Meta Ads and Shopify, and can build reports, dashboards, internal apps, marketing campaigns, code, and recurring automations across systems that previously required separate logins, separate teams, and separate budgets.

The speed of Viktor's rise reflects something deeper than clever engineering. It reflects a structural shift in how companies are beginning to think about software. For two decades, the dominant model of enterprise technology was to give every employee a login to a platform — Salesforce for sales, Workday for HR, Jira for engineering — and expect them to navigate between systems. The AI agent model replaces the login with a message. Instead of logging into five platforms to assemble a weekly performance report, an employee asks Viktor. The agent pulls the data, formats the output, and delivers it in Slack or Microsoft Teams — the same channels where colleagues already communicate.

Viktor is not another chatbot. The company is explicit about this distinction. The agent runs autonomously on a schedule, connects to company systems, builds persistent memory of how the organization operates, and delivers finished work products rather than chat replies. It can scan public Slack channels and propose workflows it can take over. In one example cited by the founders, Viktor reviewed a complex Meta Ads setup and flagged an adjustment that saves approximately $10,000 per week in advertising spending — a change a human analyst might have taken days to identify.

"We see Viktor as the first true AI employee," Wiatrowski wrote, describing the vision that has driven the company since its founding in 2023. "Not another tool. Not another tab. Not another copilot that still needs babysitting."

This positioning — an employee, not a tool — is either brilliant marketing or a genuine philosophical distinction, depending on whom you ask. But it has clearly resonated. Accel partner Zhenya Loginov told Fortune that his firm had met the founders several times over the past two years as they launched earlier products, including JaceAI, an email assistant that attempted to manage inboxes automatically. Those earlier tools impressed on technical grounds but felt too narrow to support a large company. When Viktor launched, Loginov said Accel flew to Warsaw within weeks.

The firm had been searching for a team-focused AI assistant rather than another personal helper, and Viktor's architecture — designed from the ground up to operate across an organization rather than inside a single user's workflow — matched the thesis. "Most of these products, if not all of them, are focused on individual assistance," Loginov said. "Not for the team, and that was one of the big product choices that Viktor made that we thought was fantastic."

The Warsaw Advantage

Viktor's geography tells its own story. The company is headquartered in Warsaw, with a second office in Munich and a newly opened presence in New York. It was founded not in the dense venture ecosystem of Silicon Valley but in Central Europe, where engineering talent is deep, labor costs are lower, and the pressure to compete for attention against a thousand other AI startups is less acute.

The founders met at Meta, where Albert had joined with the explicit goal of finding a co-founder. Both are Polish engineers who cut their teeth inside one of the world's largest AI organizations before deciding to build independently. Their team now includes engineers and operators with experience from Amazon, Google, Meta, and Tesla — a brain trust assembled in a city better known for its post-Soviet architecture than its startup ecosystem.

The funding round is the largest ever for a Polish-founded company, a milestone that may signal the maturation of Central Europe's technology sector. Warsaw has quietly built a reputation as an engineering hub, producing companies like ElevenLabs (synthetic voice) and DocPlanner (healthtech) that have attracted global venture capital. Viktor's round — with Accel, one of Europe's most established venture firms, in the lead — suggests that the region's startup scene is entering a new phase.

The Competitive Landscape

Viktor is entering one of the most crowded and well-funded categories in enterprise technology. Microsoft has been weaving Copilot into Teams and the broader Office suite, giving it a distribution advantage that no startup can match. Salesforce has its own Agentforce product, which builds and deploys AI agents on top of customer data and operates inside Slack — the very platform where Viktor lives. Google is integrating AI agents into Workspace. A wave of startups, from Sierra to CrewAI to Cognition, is targeting overlapping segments of the agentic AI market.

The incumbents have advantages that Viktor cannot easily replicate: existing customer relationships, massive distribution channels, and the ability to bundle AI assistants into software licenses that companies are already paying for. Microsoft can put Copilot in front of hundreds of millions of Teams users without spending a dollar on customer acquisition. Salesforce can deploy Agentforce to its installed base of 150,000-plus customers. Viktor must win each account individually.

But the incumbents also carry structural disadvantages that startups like Viktor are designed to exploit. Microsoft's Copilot is optimized for Microsoft's ecosystem; Salesforce's Agentforce is optimized for Salesforce's data. Viktor is platform-agnostic, designed to connect to over 3,000 tools regardless of vendor. It lives in Slack and Teams simultaneously, meaning a company that uses both platforms — as many enterprises do — can deploy Viktor across its entire communication stack rather than choosing one vendor's assistant.

The broader bet is that AI agents will become a new layer of enterprise infrastructure — not owned by any single platform vendor, but operating across them. If that bet is correct, the winners will be the companies that build the most capable, most integrated agents, not the companies that control the largest installed base of a particular software suite. Viktor's early traction — 2,000 organizations, $15 million in annualized revenue, a 10-week sprint to scale — suggests that at least some portion of the market agrees.

What This Signals

The $75 million round is not merely a funding event. It is a signal that the AI agent market is moving from prototypes to payroll — and that the companies that figure out how to embed agents into the existing communication fabric of the enterprise will have a structural advantage over those that require users to adopt new interfaces.

Viktor's core insight is deceptively simple: people already live in Slack and Teams. They do not want another dashboard, another login, another tab to monitor. An AI agent that lives where the conversation already happens — that can be summoned with a message rather than a platform migration — reduces the adoption friction to nearly zero. The agent becomes part of the team because it operates in the same channels as the team.

The risks are considerable. Viktor is competing against some of the largest technology companies on Earth, each of which has resources, distribution, and strategic motivation to dominate the AI agent layer. The company's $75 million war chest is substantial by startup standards but minuscule compared with the tens of billions Microsoft, Salesforce, and Google are investing in AI. The path to an independent future likely runs through a level of execution — on product, on go-to-market, on customer retention — that few startups achieve.

But the early returns are striking. A $15 million revenue run rate achieved in 10 weeks, from a standing start, is the kind of number that silences skepticism. The angel investor list — Butterfield and Henderson, who built the very platform Viktor operates inside — suggests that the people closest to the problem see the opportunity most clearly. When the co-founders of Slack write checks into an AI agent that lives inside Slack, the signal is hard to ignore.

Viktor is not the first AI agent startup to attract venture capital, and it will not be the last. But its trajectory — from a side project in a Meta AI lab to a $75 million round in roughly three years — captures something essential about the current moment. The AI agent market is forming in real time, and the companies that can deliver working products, not just compelling demos, are separating from the pack. Viktor, with 2,000 customers and a revenue curve that bends sharply upward, has placed itself in that group. The question now is whether it can stay there.