Spektr raises $20M for AI agents that handle KYC and KYB in financial compliance

Spektr wants to remove one of the most expensive and slowest parts of financial compliance: manually assembling a case file from dozens of documents, registries and internal systems. The Copenhagen startup has raised $20M in a Series A round to scale its platform of AI agents, which take over KYC and KYB checks, document analysis, ownership-structure mapping and risk-rationale preparation in minutes rather than hours.

The company is not building one universal assistant but a set of specialized agents for specific stages of the compliance process. One agent can extract data from corporate documents, another can match owners to legal entities, and a third can turn the risk-assessment logic into a clear explanation for the internal team. This approach matters especially for KYC and KYB, where finding information is not enough: it has to be verified, linked together and documented so that the decision can be defended before an auditor or regulator.

KYC is the check of an individual; KYB is the check of a legal entity. The latter is usually harder: you need to understand who actually controls the company, how the chain of owners is structured, whether it contains sanctioned or high-risk links, and whether the data in the documents matches what registries show. That is why ownership mapping and risk rationale, which Spektr has made separate platform features, matter so much to the market. Compliance teams are not satisfied with an answer of "high risk" or "low risk"; they need a transparent reasoning trail they can rely on in internal approvals and external reviews.

For banks, payment services and fintech companies, this is a painful area. When onboarding a new client, compliance staff often copy details by hand from application forms, statements, corporate registries and identity documents, then transfer them into internal systems and reports. This step wastes time, raises the cost of each check and increases the risk of human error. Spektr's promise is to leave control and the final decision with the human while removing mechanical work that adds no extra value.

The $20M Series A was led by NEA. Northzone, Seedcamp and PSV Tech also participated. For a relatively young company, this signals that investors see demand not for "yet another AI assistant" but for infrastructure embedded in the critical processes of financial institutions. Compliance is a rare segment where speed matters almost as much as accuracy: if a check takes too long, client onboarding stalls; if it is done badly, the company faces regulatory and reputational risk.

Interest in such products is logical for another reason. After the generative AI boom, many companies first experimented with chatbots and internal assistants, but they are now increasingly looking for systems that deliver a direct operational effect. In Spektr's case, that effect is easy to state in business terms: open accounts and onboard new clients faster, refresh checks on existing counterparties faster, and relieve expensive analyst teams of routine copy-paste. If the platform reliably gathers documents, understands ownership structure and explains why a given case falls into a risk zone, it is no longer a technology demo but a working tool for a key process.

The main takeaway is that AI is moving away from front-end interfaces and deeper into a business's internal operations. Spektr's story suggests that the next wave of value may be created not around general-purpose models but around narrowly specialized agents that can work by the rules, assemble verifiable context and speed up decisions in regulated industries. For fintech, this is one of the clearest ways to monetize AI right now.