The first thing you notice on Seamflow's homepage is the verb. Not "streamline," not "optimize," but "turning months into hours" [Seamflow]. It is the kind of promise that, in consumer software, would read as marketing exuberance. In the testing, inspection, and certification industry, where a single technical file for a Class IIb medical device can sit in a notified body's queue for the better part of two years, it reads as something closer to a dare.
That dare is what the London startup is being funded to make good on. Seamflow, founded in 2025, raised $4.5 million in a seed round co-led by Northzone and Initialized Capital, with participation from Nebular, Transpose Platform, Entrepreneur First, Charlie Songhurst, and the former German international footballer Mario Götze [Nordic 9, July 2025]. The company describes itself as an AI-native operating system for TIC organizations, and its opening wedge is the Notified Body Suite: software that sits across intake, technical file review, and certificate issuance for the regulators-of-regulators who decide whether a hip implant or an MRI machine can be sold in the European Union [Tely.ai, 2026].
The bet
The TIC industry is one of those quiet pieces of infrastructure that most software founders never see and most consumers never think about. Notified bodies are the private organizations that the EU delegates to assess conformity with regulations like the Medical Device Regulation (MDR). Their reviewers spend enormous amounts of time on document handling, cross-referencing, and audit trails: exactly the kind of structured-but-judgment-heavy work that current generation language models have started to handle credibly. Seamflow's pitch, articulated by co-founder Sinan Kircova, is that AI should absorb the administrative weight so that TIC experts "spend their time on judgment rather than administrative tasks" [LinkedIn - Sinan Kircova, 2026].
The wedge is MDR. According to Tech Funding News, Seamflow raised specifically to address the EU's roughly 20-month medical device certification delays [Tech Funding News, 2026]. That backlog is not theoretical. It has reshaped product roadmaps at every European medtech company since MDR came fully into force, and it has pushed some manufacturers to launch in the United States first. A piece of software that meaningfully shortens the queue is, in principle, something the notified bodies themselves want to buy.
Why it could be big
The investor list is the strongest external signal. Northzone has a long record in European software, Initialized Capital is one of the more disciplined Bay Area seed funds, and Entrepreneur First, the talent investor that originally paired the founders, returned for the round [Nordic 9, July 2025]. Charlie Songhurst, a prolific angel known for picking up unglamorous infrastructure companies early, signed on as well. The roster suggests a thesis: that the regulatory back office is large, slow, and structurally underserved by software, and that the right team can build the system of record for it.
The upside, if the thesis holds, is that Seamflow becomes vertical software for a category that almost no generalist SaaS company will ever touch. The incumbent most often named in the space, TrackWise, is a quality-management suite rather than an AI-native review engine, which gives a new entrant room to define the workflow rather than retrofit it.
Seed round (Feb 2026) | 4.5 | $M
The team and traction
Seamflow's founders are Yusufhan Kircova and Konstantin Klingler, both alumni of Entrepreneur First's London cohort [Crunchbase, 2026] [LinkedIn - Jonny Clifford, 2026]. Kircova previously worked at SAP and Bosch, two companies whose enterprise sales motions and industrial customer bases map closely onto the TIC buyer [RocketReach, 2026]. Klingler, listed as CEO, is a Schwarzman Scholar [The Org, 2026] [LinkedIn - Konstantin Klingler, 2026]. The company has grown past ten team members in its first year [LinkedIn - Yusufhan Kircova, 2026], and Pieter Verbessem is among the early hires [LinkedIn - Pieter Verbessem, 2026].
It is a small team for the size of the problem, which is precisely the shape of company that the seed investors appear to have wanted to back: domain-fluent founders, a regulator-facing wedge, and enough capital to spend the next year embedded with design partners rather than chasing logos.
The honest counterfactual
The bear case is straightforward. Notified bodies are conservative buyers. They are audited themselves, they carry legal liability for their conformity assessments, and they have historically preferred software they can fully control over AI systems whose outputs are probabilistic. A competitor like TrackWise has spent years earning trust in adjacent quality-management workflows, and any AI-native entrant has to clear a very high bar on auditability before a notified body will let model output anywhere near a certificate decision.
Seamflow's answer, visible in its own product framing, is to lean into auditability rather than around it: the Notified Body Suite is described as covering "intake to certificate issuance for TIC workflows, ensuring auditable regulatory processes" [Tely.ai, 2026]. In other words, the company is not pitching the model as the decision-maker. It is pitching the model as the assistant that prepares the file, surfaces the inconsistencies, and leaves the human reviewer with a defensible paper trail. That is the version of the product the regulators can plausibly say yes to.
What to watch
The next twelve months will be about named design partners. The TIC industry is small enough that one or two notified bodies publicly running Seamflow on live MDR files would change the conversation, and quiet enough that the absence of such announcements would also be informative. Watch for expansion beyond medical devices into adjacent regimes such as the Machinery Regulation or the AI Act's own conformity assessment requirements, which would signal that the platform is generalizing the way the investors are betting it will. A Series A in the back half of 2026, if it arrives, will be the cleanest external read on whether the months-into-hours promise is starting to hold up inside real review queues.
The cultural question Seamflow is implicitly answering is the one that has hung over enterprise AI since the first GPT-4 demos: in the parts of the economy where being wrong is genuinely expensive, will the people whose judgment the system depends on actually let the software help?