Summary

• Vessla has emerged from stealth with $3.3m in pre-seed financing led by Seedcamp, joined by Inception Fund, Tiny VC, and individual investors.
• Its AI agents are designed to work across existing property software, handling activities involving maintenance, supplier coordination, documents, and invoices.
• The company reports relationships with major Nordic property operators but has not identified those customers or published independently verified efficiency gains.

Stockholm-based Vessla has raised $3.3 million to develop AI agents that coordinate work across the software used by property owners and managers, targeting the administrative processes connecting maintenance requests, contractors, documents, and invoices.

The pre-seed financing was led by Seedcamp, with participation from Inception Fund, Tiny VC, technology industry operators, and individual investors. Vessla emerged from stealth on 19 October after developing a platform designed to operate alongside customers’ existing property management applications.

Vessla says it is already working with some of the largest commercial and residential property operators in the Nordic region. The company has not identified those customers or disclosed how many properties use its software, making it difficult to assess the scale of deployment or the benefits achieved.

The business is developing an agent-based system that retrieves information from different applications, follows customer-specific procedures, and completes defined tasks. Employees are intended to retain control over decisions involving judgement or approval, rather than delegating unrestricted authority to the software.

Property workflows rarely stay inside one application

Property management involves recurring activities that draw on information held by different teams and systems. Maintenance, leasing, financial administration, and supplier management may be handled by separate applications introduced at different stages of a property’s operation.

A routine maintenance request can illustrate the resulting complexity. A tenant may report a problem through email, while the relevant building information sits in property management software and the contractual arrangements governing repairs are recorded elsewhere.

Before work can begin, an employee may need to establish responsibility, check whether a contractor is available, obtain spending approval, and provide the supplier with accurate information. Once the repair is complete, another process may be required to reconcile the invoice against the original work order.

Those steps are not necessarily difficult individually, but managing them across applications can produce repeated data entry, fragmented records, and delays while information passes between employees.

Vessla wants its agents to operate across those boundaries, gathering context and carrying out permitted activities without forcing customers to replace the business applications in which their records are already held.

The approach differs from adding another interface that merely displays information from several systems. An agent may also execute steps in an established process, such as preparing a supplier request or recording the status of a maintenance issue.

However, the company has not released detailed independent performance measurements showing how reliably its agents complete those activities or how much employee time they save.

Reliable automation depends on local operating rules

Property operations combine repetitive activities with exceptions requiring commercial or technical judgement. A routine request for a minor repair may be processed under an established spending limit, while an emergency involving building safety can require immediate escalation and different authorisation procedures.

Agents must therefore understand more than the information appearing in an individual request. Relevant context may include the type of property, service contracts, supplier availability, budgets, and responsibilities assigned to particular employees.

Vessla says its platform applies operating rules defined for the customer, allowing agents to complete appropriate work while leaving consequential decisions to the responsible people.

That division will be important as the system develops because not every activity can be reduced to a standard sequence. A dispute concerning damage to a property, for example, may involve uncertain evidence, contractual interpretation, and competing interests.

Automating the collection of supporting information could still be valuable, but deciding who should pay for the damage may require human consideration of the evidence.

Technical integration presents another challenge. Existing property systems may expose different interfaces, apply inconsistent data structures, or contain incomplete records. An agent that depends on those systems must cope with the resulting limitations rather than assume that every record is accurate and accessible.

Permission management is equally important. Property businesses can hold personal information concerning residents, confidential commercial agreements, and financial data that should not be available to every employee or automated application.

The system’s ability to retrieve information and execute actions must therefore be limited to appropriate resources. Human approval procedures also need to establish which individuals can authorise particular activities and how those decisions are recorded.

Founders bring experience in enterprise AI

Vessla was founded in 2026 by chief executive Orman Gazvini, chief technology officer Kasper Janehag, and chief scientist Arash Dabiri.

Gazvini previously worked at McKinsey and in strategy and mergers and acquisitions at Pophouse. Janehag worked on AI programmes at QuantumBlack, McKinsey’s AI division, following earlier experience in aerospace, including work associated with NASA’s Mars 2020 mission.

Dabiri’s experience includes developing and evaluating agent-based AI applications at Andon Labs and QuantumBlack.

The investment round attracted several individual backers with experience in Nordic technology companies, including Lukas Saari of Tandem Health and Fredrik Hjelm of Voi, alongside people associated with Lovable and OpenAI.

The participation of individuals working at those businesses should not be confused with an investment by OpenAI or Lovable as companies.

Seedcamp has also added Vessla to its portfolio of 2026 investments. The funding is intended to support continued platform development and deeper connections with the systems already used in property operations.

Its ambitions extend beyond administrative maintenance work into leasing, building operations, and investment planning. Those areas involve different levels of financial responsibility, suggesting that expansion will require additional technical controls and evidence of reliable performance.

Customer results remain to be demonstrated

The company enters a market where property businesses already use established management, finance, and facilities software. Their willingness to adopt another technology layer will depend partly on whether it can work with those systems without requiring extensive replacement projects.

Retaining existing applications could reduce disruption, but it creates a continuing obligation to maintain integrations when vendors update their products or customers change their processes.

The financial benefits will also depend on the amount of work that can be completed reliably without additional employee intervention. Faster handling of an incoming request provides limited savings if another employee subsequently needs to investigate errors or repeat the process.

Vessla has not disclosed subscription pricing, recurring revenue, independently measured productivity gains, or detailed customer case studies. Its reported relationships with large Nordic property operators establish potential opportunities to test the platform, but do not yet demonstrate wide adoption.

Investors are backing the prospect that AI agents can coordinate processes across existing property technology rather than requiring every operator to undertake a major software replacement.

With the financing completed, the next commercial evidence will come from customer deployments showing whether agents can handle routine work consistently, respect local operating rules, and reduce the time employees spend moving information between disconnected systems.