Viktor raised a $75M Series A led by Accel in May 2026, with angel participation from Slack co-founders Stewart Butterfield and Cal Henderson and Vercel founder Guillermo Rauch.

Former Meta engineers Fryderyk Wiatrowski and Peter Albert developed Viktor through two earlier product iterations. At the financing, the company described three years of development by a six-engineer team drawn from Meta, Google, and Oxford, working in Warsaw and Munich.

Viktor is an AI coworker that employees use inside Slack and Microsoft Teams. A team installs it in its workspace and connects business software through more than 3,200 integrations, giving it access to the tools and data needed for the work it assigns.

For a weekly advertising audit, a marketing manager can ask Viktor to pull campaign spending and results from connected ad accounts and produce recommendations. Viktor analyzes the data in its cloud workspace and returns a formatted report in chat, where the manager reviews the findings before authorizing changes.

The persistent cloud workspace lets Viktor write and run code, retain context from prior work, and schedule recurring jobs. Per-person access controls and approval prompts govern what it can read or change, allowing the same workflow to combine analysis with actions across connected systems.

Viktor charges for shared workspace credits rather than seats, with its Team plan priced at $100 per month for 40,000 credits. Enterprise customers negotiate volume pricing and billing terms, while adding employees to a workspace expands the pool of potential users without creating another seat charge.

Credits map to model work performed across providers such as Anthropic and OpenAI, so a customer's bill depends on the amount and complexity of work delegated. More recurring workflows can raise account spending, while caching reduces the model usage and credits required to repeat a task.

Viktor's economic value capture depends on the difference between what it charges for credits and its actual inference and execution costs. Lower provider costs could improve that spread, but caching that also lowers customer consumption or enterprise discounts that reduce the price per credit can pass some of the benefit back to customers.

A larger volume of credit billings therefore does not establish an equivalent improvement in gross profit. The credit prices alone do not reveal that spread, and Viktor must also fund the integration maintenance and support needed to make a broad AI coworker useful across different software stacks.

Viktor competes with Microsoft's Copilot agents, Salesforce's Agentforce in Slack, and independent cross-application automation providers such as Zapier. The central competition is between agents distributed by enterprise platform owners and a packaged coworker that works across a customer's existing mix of software.

Microsoft can distribute agents through Teams and Microsoft 365, while Salesforce brings Agentforce into Slack using existing Salesforce and Slack permissions. Both already have enterprise customer relationships, administrator controls, and access to data in their own systems, giving them an advantage when an organization evaluates another agent vendor.

Viktor offers an independent execution layer for a team using, for example, HubSpot, Linear, and Google Drive together. Its commercial case depends on making that cross-tool work easier to delegate than building and governing a custom agent within an incumbent platform, rather than merely offering another chat interface.

Zapier is integrating agent functionality into its existing automation platform, where customers can combine AI steps with established triggers, actions, filters, and automation history. Its existing users already have workflows and application connections that can support adoption of new agent capabilities.

Viktor packages execution as a coworker inside the team's chat workspace, reducing how much workflow design users need to do before assigning a task. It has to deliver reliable results across connected systems while preserving enough user control to compete with an automation product whose steps and execution history teams can inspect.

Viktor could enter managed-operations budgets by taking responsibility for complete business processes, such as recurring billing reconciliation, rather than supplying employees with a tool for individual tasks. Its connected execution workspace provides part of that infrastructure, but entry would require process-specific exception handling, audit trails, and contractual responsibility for outcomes.

Platform competition: Viktor sells an additional workspace product inside the collaboration channels where Microsoft and Salesforce distribute their own agents. It must make independent cross-tool execution valuable enough to justify another vendor and budget alongside those existing enterprise relationships.

Margin capture: Viktor must keep the cost of delegated work attractive to customers while retaining money above its inference and execution costs. Enterprise volume discounts and caching that lowers customer credit consumption can improve adoption without guaranteeing a corresponding improvement in its own gross profit.

Review burden: Viktor's broad remit puts advertising accounts, customer records, and deployed applications within the scope of one coworker's actions. If teams must reconstruct each multi-step workflow before approving it, the review labor can absorb the operational savings they expected from delegating the work.