From first build to repeatable production.
I help robotics, autonomy and deep-tech hardware startups turn a prototype that works into a product they can build, ship and scale, without losing the speed that got them there.
The problems that show up between prototype and product
Your second build takes as long as the first
The prototype works, but nobody can build it again without the person who built it.
The plan lives in one founder's head
It worked at ten people. At thirty, decisions stall and the same arguments keep coming back.
Sales promises dates engineering can't hit
Customers are easy to win and hard to keep, because commitments outrun what the lab can deliver.
You've raised the money; now you have to build
Suppliers, tooling, hiring and certification all land at once, with investors watching the milestones.
Start with a free review. Then only what you need.
Scale-up review
A 60-minute session on your plan, team and next milestone. You get a one-page summary of the risks I'd tackle first, whether or not we work together.
Interim program leadership
For a milestone that needs someone full-time: building the plan, running the program and the cadence, and handing it over to your team.
Targeted support
For sporadic needs: design and gate reviews, supplier and build planning, investor and board preparation, scale-up cost models.
Scope and price agreed together after the review.
I've been inside the transition, more than once
Joined at around 100 people and left at around 1,000, through the initial start of production. [CONFIRM: "4 VPs built" — wording to be agreed.]
Defined and drove the program through two key milestones, to the point where the company was successfully acquired.
The company grew from 20 to more than 60 people and closed a successful Series B.
Automotive engineering and programs at Jaguar Land Rover and McLaren, where I learned what production discipline looks like at scale.
What it takes to scale a delivery-drone fleet
Using only published information about an Indian drone-delivery startup, I modelled how build cost, factory staffing and cost per delivery change as the fleet grows. It shows the kind of analysis a scale-up review starts with.
Build cost per drone, hand-built versus designed for volume.
Factory staff needed at 100 drones a day, for the same two designs.
Once the drone is cheap, supervision and handling decide the cost per delivery.
Independent analysis from public sources; not a client engagement.
“[RECOMMENDATION TEXT — paste the strongest lines from David Hsu's LinkedIn recommendation.]”
I'm a program leader and mechanical engineer with more than 15 years in automotive, autonomous vehicles and AI hardware. I've worked from large manufacturers to ten-person startups, and the problems I enjoy most sit in between: the moment a company has to stop building one of something and start building many.
I'm based in Denmark and work with startups across Europe, on site or remotely.
Denmark
English (native), Spanish (intermediate)
Robotics, autonomy, drones, deep tech