Stan Callewaert

The through-line

Stan Callewaert's consistent argument is that voice is the fastest way to get information out of a clinician's head and into software, and that the whole business is a race to make that obvious to Europe before someone else does. He frames Squire's position as "a blue ocean, certainly in Europe," noting that "similar solutions exist in America and so on, but in Europe it's really an untapped market and I think it's going to become very competitive, so speed will be very important to us" [2]. What sits underneath that urgency is an unusually patient product method: deliberately small, deliberately hands-on, deliberately slow to scale, on the view that the depth of the early iterations is what makes the later speed possible [2].

The arc, so far as the material shows it, runs from a GP-facing transcription tool built invite-only with individual doctors [2] to embedding inside the software an entire national home nursing organisation already uses, with Squire integrated into Wit-Gele Kruis [1]. The ambition attached is explicitly continental: "making impact on the daily lives of thousands of home nurses all over Europe" [1].

On why voice beats typing

The case for the product is arithmetic, and Callewaert made a point of measuring it himself rather than asserting it. Following a shift with home nurses, he "clocked their typing at about 30 words per minute. When speaking, they can hit up to 150 words per minute" [1]. That five-fold gap is the entire value proposition, and he describes the alternative in terms of friction rather than features: instead of "wrestling with dropdown menus, date pickers and typing out text fields, they can just say the information and everything gets filled in automatically" [1]. The unit of progress is the clinician's time, and the claim he makes for it is blunt: "This saves a massive amount of time" [1].

On fitting into software people already use

The Wit-Gele Kruis integration is instructive about how he thinks distribution works in healthcare: home nurses "can now simply talk to their tablet to fill in patient information directly in the software they were already using" [1]. Nothing is ripped out and nothing is replaced. Squire attaches itself to the incumbent workflow rather than asking a nurse mid-shift to adopt a new one, and the win is framed as reaching the biggest Belgian home nursing organization rather than as a headline user count [1].

On staying invite-only and iterating one GP at a time

Against the instinct to chase growth, Squire has kept the product invite-only and avoided chasing big user numbers, meeting each GP one-on-one and returning every three weeks to iterate on product feedback before scaling [2]. That answers the product-led versus sales-led question in a specific way: the early motion is high-touch and manual by design, and volume is treated as something to earn later rather than something to optimise for now [2]. It also sits in productive tension with the speed argument, and the resolution is that time spent close to a small number of users is what makes the eventual land-grab worth winning [2].

On hardware as a temporary crutch

Callewaert treats the microphone as a phase, not a business. Hardware matters early because it improves the quality of iteration, but he expects it to become irrelevant over time as Squire becomes a pure software product, and Squire does not build mics itself, which keeps scalability intact [2]. The principle is that you may lean on a physical component to learn faster, provided you never take on the manufacturing burden that would cap how far the software can travel [2].

On competition and European timing

The strategic read is that America already has comparable solutions and Europe does not, that the gap will close, and that being early is only an advantage if it is converted quickly [2]. He expects the market to "become very competitive," which is why "speed will be very important to us" [2]. The framing of the Wit-Gele Kruis launch matches this posture: it is described as "only the beginning," with the next horizon being thousands of home nurses across Europe [1].

On credit and how the work gets done

Even the wins are described as collective. The integration is attributed by name to Edward Maddens, Korneel Traen, Kristof D. and Nelis Goeminne "and many others," as "the people making this happen" [1], and the founder's own contribution he chooses to highlight is the shift he spent alongside the nurses with a stopwatch [1].

Takeaways

• Measure the problem in the field before selling the solution: nurses type at about 30 words per minute and speak at up to 150, and that gap is the product [1].
• Europe is "really an untapped market" for AI medical transcription while America is not, so the window is real but temporary and "speed will be very important" [2].
• Deliberately staying invite-only, meeting each GP one-on-one and iterating every three weeks, beats optimising for user numbers before the product is right [2].
• Use hardware to sharpen early iteration, but do not build it: keeping the microphone out of scope is what keeps the company a scalable software business [2].
• Win distribution by embedding in the systems clinicians already open, so voice replaces dropdown menus and date pickers inside existing software rather than adding another tool [1].
• Treat a national rollout as a starting line, not a finish: the stated goal is thousands of home nurses across Europe [1].