RimaruBook a call

What I built, did and studied

Things I've built for clients, contract work, client projects and research.

Built for a client

Harvest Network marketplace

I saw someone looking for a team to build a flower marketplace, reached out, won them as a client and delivered it.

Someone wanted a marketplace for flowers, farmers and wholesale buyers. Nobody had asked me. They had no team, so they needed someone to scope it and get it built.

I scoped it so we both knew what was in and what was out, put a team together, ran the build and delivered the whole marketplace.

Team project. I led it.

Contract work

Rogers innovation consulting

Rogers brought me in on a contract to work out what else it could do beyond connectivity.

I can't say what the problem was or what came out of it, because that's Rogers' to share. What I can tell you is the kind of work it was and what I did.

It went the way most consulting work goes: research, then ideas, then proof, then a deck. I wrote the hypotheses, designed the interview guide and the survey, ran the interviews, and built the deck.

Innovation consultant

Client projects

Canada Post consulting project

Canada Post was a client of our team, and I was the person on our side they talked to.

It was a consulting project, and we worked with them directly. I was the client liaison and the interviewer, and I ran the sessions with them. I can't say what the project was about or what we found, because that's Canada Post's to tell.

Team project. I led the client sessions.

Research projects

SMIC under US-China decoupling

This was research. We picked the company to study, and we never worked for it.

We looked at how China's largest chip foundry keeps innovating under US export controls, using only public sources.

SMIC can't buy the extreme ultraviolet machines that print the smallest chips, because US export controls cut it off from the best chip-making tools. So it works on two timelines: earn money now, reach advanced chips later.

Where we landed: the older, mature chips in cars and home electronics pay the bills, and that profit funds the push toward the advanced ones.

Team project.

Want to tell me what challenges you're facing?

Kitchener challenge marketplace

Solo project. I worked alone.

The question

The City of Kitchener had run a pilot to connect city problems with outside innovators. I was asked to pick one of its three parts and write an implementation concept for an early rollout. I picked the challenge marketplace, because it suits problems with a lot of people involved.

The real problem underneath

The pilot proved the idea, but it left some gaps. Nobody inside the city had committed to owning a challenge. The intake was complicated. Vendors mixed up pitching with procurement. And there was no path from a good idea to a decision. The lab is small, so a bigger funnel wasn't the answer. It needed one way in that a small team could run the same way every time. Most of my design went into what happens before a challenge goes public.

What I did

  • I looked for the closest versions of the problem. I searched worldwide, and asked four things of each one: what did they build, did it work, what broke, and what did they get.
  • I picked the best match. Ukraine's Brave1 defence marketplace.
  • I read the city side. Published city documents and regional partner pages.
  • I used AI for research and drafting. The concept says so. The design calls were mine.

What I found

  • Brave1 flipped who writes the requirement. The team with the problem writes it in structured fields, vetted builders respond, and a separate agency handles contract and payment.
  • Most submissions describe a solution, not a problem. "We need cameras" often hides a cause like thin supervision. A good statement names a cause, names no supplier, is one problem, and says how success will be measured.
  • The usual measures count activity, and that doesn't show whether it worked. Outcomes do, and they look worst in the first year.

What I recommended

  • Rubric first. Agree what a good challenge contains, then build the form from it.
  • Two gates. The problem needs data behind it. Then a division director commits staff time in writing before anything goes public.
  • Three phases. A submission can end five ways: advance, redirect, hold, close with a reason, or go to procurement.
  • AI does the sorting. People make the decisions.
  • A 12-week pilot on tools the city already had, with no new budget and stop signals set in advance.

What I'd do next

  • I'd run the rubric on a handful of real staff problems first, to see whether it turns "we need cameras" into a cause.
  • I'd time the director gate in the pilot. If shaping one challenge takes too long, the gate is too heavy.
  • I'd ask the city how much of it they want automated. I kept most steps human, so they can decide.