Programmify

Started

2023

Status

Live

Signal

Education

Open product

Why it exists

I learned to build the slow way: scattered tutorials, no one to tell me which parts mattered, and long stretches of working alone without knowing whether I was getting better. Years later I was on the other side of it, teaching computer science on hourly pay at a university and freelancing as a web developer at night. My students were writing code on paper and on whiteboards, the same thing I hated when I was the one in the seat. When internship season came, the only placements open to them were in offices with no engineering work in them. There were not enough tech companies willing to take them.

So in 2023 I ran my own version instead: a startup-style internship where the work was real. Four students. Then seven. Those seven became the first co-mentors, and with them we ran the third edition and it stopped being a university thing. 108 interns, from Nigeria, Kenya, South Africa, Ghana and Turkey. No marketing budget, no connections, no experience running a company. The fourth edition took 300.

The line we use is “from learner to founder,” and it describes a route, not a slogan. Put people in cohorts, give them real work instead of a syllabus, and make shipping the thing you are graded on.

Where it is now

400+ builders trained, across 15+ countries, with 100+ projects shipped. A community of 1,240+ follows the work on LinkedIn, and a good share of them are people who came through a cohort and stayed.

The ecosystem is a pipeline, and each stage assumes the one before it.

• Training Program. The entry point. Fundamentals, then AI-augmented work: prompting, agents, and shipping something real.
• Builder Internship (PIP). The Programmify Internship Program puts people into AI-augmented squads on actual project work.
• Founder Fellowship. An eight-week accelerator for people who have a product and need to turn it into a company.
• Hackerhouse. An immersive live-in cohort. Everyone in one place, building, for a fixed stretch.

The AI-native turn

We used to teach programming. Now we teach people to build with agents. The academy tracks are organised around AI-first startups, AI-augmented squads, and agentic AI, and the pitch is that we are pioneering the first AI builder lab in Africa.

That shift was not a rebrand. The work students were producing changed first, and the curriculum caught up.

We also run recurring 48-hour AI Weekend Build hackathons in partnership with Lovable. Selected participants get free Lovable Pro access for the weekend, and winners take up to 10,000 Lovable credits along with mentorship and a startup pack. It is the cheapest way we have found to show someone what they are capable of in two days.

Why Ubuntu is not decoration

“I am because we are.” It is on the site and it is the reason the programs are cohort-based rather than self-serve. People learn faster with witnesses. A builder who ships in front of fifteen peers behaves differently from one grinding alone through a video course.

My part in it

Co-founder. In practice that has meant curriculum design, teaching, mentoring and product decisions, and increasingly the argument about what a program should look like when the tools write half the code.

What I learned

Teaching made me a better builder. Explaining something shows you fast which ideas only sound simple because you already understand them, and confusion turns out to have a repeatable shape you can design around.

The other lesson is that learning software is emotional and social as much as technical. People need permission to be beginners and enough pressure to keep moving. That balance shapes how I think about PeerDrift, FastCV and OpenBuildLab.