Emergency physician with no coding background built six AI products serving 1,100+ delegates

At a glance: 2 conferences powered · 1,100+ delegates served · 2 institutional recognitions in writing · 6 products built · ~$717 total tooling cost · $0 capital raised · Built by a physician with no coding background

The problem I started with. I am an emergency physician in Tanzania. A career in medicine here means living inside systems that are still being built. Emergency medicine itself is a young specialty in my country, and I was part of establishing it. The gaps I have felt as a doctor, a researcher, a mentor, and a mother are not failures; they are unfinished work. I came into AI Mastery carrying a list of them, and I started with the two I knew best. The first is small but universal: medical conferences across East Africa run on spreadsheets, email chains, and paper. Abstracts get lost, and organizers drown in manual coordination. The second is the one that costs lives: Tanzania does not yet have a formal EMS system. It is still being built. Prehospital care is the most untapped layer of our health system, and the one where coordination and knowledge could save the most lives, soonest. I have treated patients who arrived too late, not because medicine failed them, but because coordination did. I began with the conference platform because it was the one I could prove fastest, and it became the foundation for everything after.

Achievements

Kongamano (conference platform), deployed and institutionally recognized:

- Powered two real medical conferences serving 1,100+ delegates: the Aga Khan Scientific Conference 2026 (~400 delegates) and the 14th MUHAS Scientific Conference (~700 to 750 delegates).

- Both institutions put their appreciation in writing. MUHAS (Muhimbili University of Health and Allied Sciences) presented me with an innovation award for the app. They had been trying to get something like it for years without success, and I delivered it in a fraction of that time; the official Certificate of Appreciation reads "for developing and supporting the use of the MUHAS Scientific Conference mobile application." The Aga Khan Organizing Committee sent a formal letter, signed by their Head of Research, praising the technical expertise and responsiveness and noting the app made it easier for delegates to navigate and engage.

- Delegates found everything in one place: abstracts, speakers, and schedules, with an AI assistant for support and personalized agendas matched to their interests. Organizers also cut costs by eliminating printed program books, supporting AKU's net-zero mission.

- Total tooling cost: ~$717. Capital raised: $0. That is under $1 per delegate served. And while that is a small budget for what was built, it was not small for me: I paid for these tools, on top of the program itself, while unemployed and without an income. Every subscription was part of the same bet.

The Emergency Coordination Platform, the impact multiplier Kongamano unlocked:

- Kongamano proved I could build; the coordination platform is why I build. It connects what already exists in a fragmented system (the public, trained first-aid responders, ambulances, hospitals, and government) into one coordinated emergency response layer, rather than waiting decades for a formal EMS system to be built from scratch.

- With no capital, just an idea and a platform I built myself, it earned me a seat in rooms I could never have entered before: partnership discussions for a pilot in one region of Tanzania.

- The stakes, in evidence: Tanzania has one of the highest road traffic death rates in sub-Saharan Africa, with WHO-cited figures as high as 31 deaths per 100,000 people, nearly 1.7 times the global rate. Research estimates that up to 54% of deaths in low- and middle-income countries could be addressed by better prehospital and emergency care. For a pilot region of around one million people, that translates to roughly 300+ road traffic deaths per year, before counting obstetric, cardiac, and pediatric emergencies. If coordinated response averted even a conservative 10 to 20% of trauma deaths (well below what the literature suggests organized systems achieve), that is dozens of lives every year, in one region, from road trauma alone. Scaled nationally, the potential is in the hundreds of lives annually, and the disability averted is larger still. (Sources: Thind et al., Disease Control Priorities 3rd ed., 2015; WHO/UN Road Safety Fund, 2023.)

- I state this carefully because I am a physician: these are evidence-based estimates of the opportunity, not claims of results. The results come next; that is what the pilot is for. But this is what "impact" means where I live: not an app metric, but a mother who reaches a hospital in time.

- Alongside it, I built a community life-saving skills platform, because I have treated too many patients whose bystanders didn't know what to do. It teaches first aid in schools and communities, started as the smallest possible version and improved daily through feedback. Together, the two are prevention and response: a community that knows what to do in an emergency, and a system that answers when they call. Also includes emergency numbers directory for Dar es salaam in one platform.

The ripple effects, every product born from a problem I have lived:

- Ascend MD, an AI research mentorship platform for African medical students and faculty. Research is one of our continent's widest gaps. Africa carries a heavy share of the world's disease burden yet produces only a small fraction of its research, and mentors are painfully few. Ascend MD is that mentorship made scalable, coaching clinicians through designing, conducting, and writing up research faster, so we can generate the evidence that improves clinical outcomes for our own patients. I am now working through registration so universities can adopt it for their postgraduate programs.

- "The Hardest Shift", a complete book manuscript with companion AI-enhanced pumping app for new mothers in high-stakes careers, born from having my first child at 39 while chairing an emergency department, with its cover and visual identity created in Claude Design.

- Akiba Yangu, an AI-enabled financial coach in pilot, helping individuals plan, understand finance, and reach the goals that let them live their dreams, because I know firsthand what it is to chase financial freedom and need a guide.

I am not building for a market I studied. I am building for a life I have lived: solutions for us, by one of us who has seen the problems firsthand.

Tools Claude (my architect, coder, debugger, and thinking partner throughout, introduced to me by this program), GitHub (first version hosting), Airtable (v1 backend, learned in Vyk's class), Lovable + Supabase (multi-tenant rebuild), Claude Design (book and app visual identity). Next: Claude Code, using the brownfield workflow from Vyk's guide, to extend what I've already built.

Methodology Vibe-coding, run on the OODA loop (Observe, Orient, Decide, Act), which Vishen taught in class and Vyk kept emphasizing as the way to create momentum. It became my operating rhythm: observe what delegates and organizers actually struggle with, orient using Claude as my thinking partner, decide on the smallest useful build, act, then loop again. In practice: Claude wrote the code and produced the first HTML app; I tested, refined, and repeated (build, test, edit), then hosted on GitHub. When the program introduced Airtable, I connected it as the backend to collect delegate data and feedback, and used Claude to build an AI chat assistant wired into the app. That version ran my first conference of 400 delegates. As my skills grew with the curriculum, I rebuilt the platform properly as a multi-tenant application on Lovable with Supabase. Every product since has followed the same loop, block by block, the way I once learned medicine: start small, get feedback, improve every day. Momentum over perfection.

Learnings

- My formula turned out to be simple: real problems I have lived, solved with tools I learned in AI Mastery. A clinician with no coding background can ship production software that institutions formally recognize, once the right tools meet the right problem.

- A deployed, imperfect product beats a perfect plan: my v1 ran a 400-delegate conference and taught me more than any tutorial could.

- A working product is leverage. It opened doors that credentials alone never did, including rooms where a regional emergency-services pilot is being discussed.

- Life happens mid-build. My young son was admitted to the ICU during the program; I fell behind, came back, and kept building. The measure of commitment is coming back.

Who it's for, and beyond the program Kongamano serves conference organizers of medical and academic events in East Africa, with further medical conferences already in the pipeline, including events where I serve on the organizing committee. The coordination platform is built for people for whom calling for help does not yet bring help. It begins the way every honest system does: with a pilot, in one region of Tanzania, which is what the current partnership discussions are working toward. If the model proves itself there, the ambition is national, and one day, similar settings across Africa. This is just the beginning: I am still working through registration and monetization, and I haven't made money yet. I won't pretend otherwise. But everything is deployed or in motion, the recognitions are on the wall, the partnership rooms are open, and I believe I am on the right track. Check with me in a year! This is not a course project. It is the work of my life, finally buildable.

A note on links and materials: Kongamano's conference deployments are evidenced by the attached institutional letter, award, and screenshots. Some of my other platforms, including the emergency coordination platform, are not publicly linked here by design: they are subject to ongoing partnership discussions with institutional and government stakeholders, and I am keeping them confidential until agreements are in place. I have attached a single representative image for that platform, and I would be glad to give the review team a private demonstration on request. The other products pictured are built and functional but pre-launch: in pilot, gathering feedback, with formal registration in progress. They are shown as evidence of capability, not as launched businesses. I kindly ask the review team to maintain the confidentiality of these pre-launch and partnership-stage materials. Thank you.