CompanyActive2015 to now
SENPAI
A design and technology company, built in public
Senpai designs, builds and pilots the digital systems institutions depend on, and trains the people who keep them running. This is the company itself: what we are building, and every decision that shapes it, published as it happens.
- Type
- Company
- For
- Self-commissioned
- Started
- 1 December 2015
- Running
- In build
- What we do
- Systems design, Product design, Engineering, Programmes
What it is
A design and technology company in Abuja. We diagnose how an organisation actually runs, design and build the system it should run on, pilot it in the field until it works, and train the people who will keep it running after we leave. The Senpai Collective is the bench that does the work.
The problem
Africa is not short of talent, ideas, or problems worth solving. It is short of institutions that turn those into things that keep working after launch, and short of people trained to keep them running. Most digitisation here fails the same way: somebody builds it, hands it over and leaves.
Who it is for
Two groups, and we are only doing our job if both are served. Institutions carrying systems that people depend on and cannot opt out of. And the creative and technical talent of this country, who currently have no institution behind them that wants them to win.
How we are going about it
- One team for the whole arc, from diagnosis through build to pilot.
- The deliverable is a running system, not a recommendation.
- Clients own the code, designs, data and accounts from day one.
- We build the people too, through the Collective, so capacity outlives the contract.
- Every decision gets published with its reasoning, before the outcome is known.
Where it stands
Relaunched in 2026 as SENPAI FUTURES LTD. The Collective platform is live and its founding cohort is forming. Our own client intake, evidence and delivery systems mostly do not exist yet.
What is here
Field notes
In order. Start at 01 if you want the whole thing.
- 01What is a company?In 2021 I had two thousand members, four cohorts, revenue and graduates earning in dollars. I did not have a company. I have been trying to work out what the difference actually is, because we are now building the thing I did not have, in public.
- 02The company systems indexA company is the part of the work that keeps happening without you. To find out whether we have one, I listed the machinery: 41 capabilities across eight domains, and what each one has to do. Senpai scores 29 out of 100.
- 03Building a company operating systemLast note I scored us at 29 and said Demand was the constraint. So I built the demand system. Here is what I looked at first, the documents it came from, the order it got built in, and what it actually does. It is live, the score moved to 31, and you can log into it.
- 04How does an institution remember?Last note I was building the company operating system and the first pieces work. They are also empty. This is the problem underneath that: almost no company has one canonical list of the people it knows, and when a person leaves, the relationships leave with them.
- 05The Nigerian startup ecosystem mapA public, queryable map of who funds, trains, houses and builds for startups in Nigeria, city by city, maintained by those organisations themselves. Where the idea came from, the five things we considered building instead, what already exists elsewhere, and what would make us stop.
- 06What we found building the Nigerian ecosystem recordThe Nigerian startup ecosystem record is live with 110 organisations across 34 cities. Why it has to be a source of truth rather than another index, the three domains taken over by gambling sites, the design I got wrong four times, and the problem nobody has solved: how a record like this stays true after the people who built it get busy.
- 07Can a stranger tell what your organisation does?Every entry in the Nigerian startup ecosystem record now shows the sentence it came from, the page it was read on and the date. What reading 109 organisations' own websites turned up: five compromised sites, four dead addresses, and an ecosystem that runs programmes and rarely writes cheques. Plus the four things I got wrong doing it.
- 08A health ID from birth, and every trade-off behind itAn experimental answer to a NITDA Builders Festival challenge: why a "health BVN" on its own would not work, why we built on the NIN instead of competing with it, why the ID is issued at birth or at the first vaccination, how bulk-issued IDs make offline registration possible, the portable summary that travels with the ID, how matching, consent and sync are designed, the bugs that taught us the most, and what is still not built.
Follow SENPAI
New notes on this project as they are written, and everything else we publish while building the digital infrastructure Africa needs. Roughly weekly.
No pitch, and one click to leave.
A running record, so it is never finished. Notes are added as decisions get made, not once they have worked.