Current investment focus: early-stage B2B and fintech in the Benelux, and Pakistani B2B products expanding internationally. Earlier articles remain dated analysis and may cover broader topics.
A brain-computer interface startup asks investors to think on two timescales. There is the immediate technical milestone: can a system perform a clearly defined task? Then there is the longer journey towards something a person can use reliably in daily life. I would want a funding plan that connects those two questions without pretending they are the same.
For the Benelux, this is an interesting corner of AI and deep technology because the commercial discussion begins with a very specific human need. It also demands restraint. A financing announcement does not establish clinical effectiveness, market authorisation or an available treatment. This article considers business-building questions, not medical advice or an assessment of a device’s safety.
The Dutch funding signal
On 15 September, LUMO Labs announced UCANACT’s €800,000 pre-seed round, led with Utrecht Holdings Seed Fund. The Dutch startup is developing brain-computer communication technology for people with motor disabilities. This is development funding, not evidence that a finished product is available.
My interest is in how an early team turns a scientific capability into a dependable product. The useful investor conversation should make each remaining uncertainty visible. Instead of asking whether neurotechnology is a large market in the abstract, I would ask which uncertainty this particular round is intended to resolve and who will independently assess the result.
Start with the person using the system
I would begin with a narrowly described use case. Who is the intended user, what communication task matters to that person, and where would the system be used? A demonstration of a technical capability is not enough to answer those questions. Founders should explain how intended users and relevant specialists influence the design, while protecting participants’ privacy.
For a hypothetical communication product, I would want the evaluation to include the effort needed to start a session, correct an error and finish a task. Those are questions for properly designed studies, not claims about UCANACT’s performance. A system that works under one carefully managed condition still needs evidence before broader promises become credible.
The surrounding experience matters too. Who helps with setup? What happens when a session cannot continue? Can the person clearly understand what the system is doing? I would ask the team to map the complete experience, including support and interruption, rather than showing only the successful output at the end of a demonstration.
Separate technical, clinical and commercial milestones
A useful development plan would distinguish technical progress from evidence about use and from commercial access. Each can require different people, budgets and decision criteria. I would avoid one broad milestone such as “launch the platform” when several unresolved steps sit underneath it. Investors need to understand the dependencies before discussing the next financing round.
My preferred milestone is a question with a defined method of answering it. For example: can a specified task be completed under the intended evaluation conditions, with the required support documented? The team and qualified experts should determine the appropriate protocol and measures. The investment discussion should then explain what a positive, mixed or negative result changes.
There should also be a record of what remains unknown. If a presentation moves between laboratory evidence, user feedback and future expectations, label each clearly. That discipline makes a founder easier to trust. It also helps an investor avoid turning an encouraging early result into a claim the company itself has not established.
Give AI a precise responsibility
In an AI-enabled assistive product, I would ask which part of the workflow the model handles and how its output is evaluated. The phrase “powered by AI” is too broad to be useful. Explain what the system receives, what it produces, how uncertainty appears to the user and who decides whether a version is ready for further evaluation.
I would pay particular attention to the difference between intended communication and system-generated assistance. If software suggests or completes an output, how is the user’s intention preserved and confirmed? That is a product-design question I would want the team to address explicitly. It should not be hidden behind an attractive demonstration of fluent language.
Model changes also need a clear process. Ask how versions are tracked, how a change is tested and how earlier results remain interpretable. The objective is not to prevent improvement. It is to make sure the evidence describes the version being discussed, rather than a moving collection of prototypes with different behaviour.
Understand who pays and who provides support
The user, purchaser and organisation supporting deployment may be different parties. I would ask founders to identify each one rather than treating positive user feedback as a purchase commitment. Market-access work should establish the actual route for the intended product; it should not assume that interest automatically creates a reimbursed or approved commercial offering.
A credible cost model should include more than hardware and software. Training, maintenance, specialist input and ongoing support belong in the discussion. For a research-led company, the transition from a founder-supported demonstration to repeatable delivery is an important business question. It should be budgeted rather than deferred until after the next fundraising announcement.
Across Belgium, the Netherlands and Luxembourg, I would approach expansion as a sequence of validated local routes. A successful relationship in one setting is a starting point, not proof of access everywhere. Identify the next partner’s role and the evidence it needs before building a forecast around geographic expansion.
The investment conversation I would welcome
A strong first discussion would bring together the intended use, the evidence collected, the remaining uncertainties and the next funded milestone. I would also want to understand who on the team owns the scientific, product and commercial decisions. Clear responsibilities are especially useful when progress depends on several specialist disciplines working together.
My approach to early-stage investing in the Benelux is to connect ambitious technology with practical execution. If you are building assistive technology or a research-based AI company, share the problem, your evidence and the next question you need to answer. The opportunity deserves curiosity, but the company deserves a financing conversation grounded in what is known today.