Ten pitches, two winners: inside the AI Academy Kenya Pitchathon

The AI Academy Kenya is Meta's programme to put practical AI skills in the hands of Kenyan students, professionals, small businesses and startups, delivered with This Is Digital. It has two parts: a Pitchathon for startups and developers already building with Meta's AI technologies, and an eight-week Masterclass for 300 Kenyans that starts in November. The Pitchathon came first, and on Friday 16 October it brought ten finalists, a five-person judging panel and a full room to the Garage at The Social House in Lavington.

From 426 applications to ten finalists

Each eligible application was read and scored against the seven weighted criteria published on the Pitchathon page: the problem and its users, the solution and its use of Meta AI, traction and stage, market and business model, social and economic impact, the team, and pitch readiness. The ten highest-scoring applications made the shortlist, announced on 8 October. The sectors said a lot about where Kenyan builders see AI earning its keep: health, education, legal services, tax compliance, SME operations and accessibility.

The morning: why Kenya, why open-source

The day opened with remarks from Meta and from the State Department for ICT and the Digital Economy, then a fireside chat between Grace Murugi, co-founder of This Is Digital, and Chinny Francis, Meta's Policy Programs Manager for AI & Wearables, on AI skilling and the partnership behind the Academy.

Three threads ran through it. Why Kenya: one of Africa's most dynamic innovation ecosystems, a deep developer talent pool, a strong university pipeline and a government that has made AI a national priority. Why open-source: Llama lets Kenyan developers build, fine-tune and own solutions in Kiswahili and local contexts without starting from scratch, and most of the finalists were combining Llama with WhatsApp or PyTorch, tools people already use, at local cost. And why skills alone are not enough: founders need capital, visibility and access to markets and partners, which is exactly what the Pitchathon set out to combine. The Academy is already running in Nigeria, is now in Kenya, and South Africa is next.

Then the judges were introduced: Chinny Francis (Policy Programs Manager, AI & Wearables, Meta), Gregory Wanjama (Co-Founder & CEO, This Is Digital), Isaac Hunja (Founder & CEO, Kaara Works), Barnabas Sang (Director, ICT Field Services, State Department for ICT and the Digital Economy) and Natasha Nalyaka (Data Scientist and Lecturer, @iLabAfrica, Strathmore University).

Ten pitches, five minutes each

Every finalist had five minutes to pitch and three minutes of questions from the panel. In pitch order:

1. Solivia Data — AfyaEdge, an offline-first, solar-powered triage hub that replaces paper registers in clinics, with on-device AI flagging high-risk patients.
2. Neural Labs Africa — NeuralSight, an infrastructure-first tele-radiology platform that connects existing imaging equipment to remote specialists, with Edge AI and Llama drafting the reports.
3. HakiChain — an AI legal suite for African law: citation-grounded research, document generation and review, and HakiScribe, which transcribes Kiswahili–English meetings into billable hours and drafts.
4. John Digital Solutions — Sawazisha AI, a WhatsApp assistant that keeps small businesses eTIMS-ready from forwarded M-Pesa receipts, voice notes and invoice photos in Swahili, Sheng or English.
5. Me-Nasoma — a read-aloud and study companion that turns documents and web pages into audio on mobile, web and a Chrome extension, priced in Kenyan shillings and built for local languages.
6. JuaFlow — automation for the workflows between WhatsApp, other channels and internal business systems, cutting order processing from hours to minutes.
7. Jackfruit Network — the School Needs Recommender, an AI agent delivered over WhatsApp that gives low-fee private schools prioritised improvement plans and a route to funding.
8. Tembotrade — one platform for SME sales, inventory and M-Pesa payments, with a Llama-powered assistant running WhatsApp commerce in Swahili and Sheng.
9. O.C.A. (Offline Claims Audit) — an offline, edge-based engine that audits SHA medical claims on Android tablets before submission, catching clerical errors and code mismatches.
10. Utter Africa — an affordable, browser-based communication platform for early-verbal and non-verbal children, speaking Kiswahili and local Kenyan languages from tapped symbols.

The judges scored each pitch on five things: how effectively it used Meta's AI technologies, how clearly it solved a local problem, its market potential and scalability, the state of the product, and the strength of the presentation and team.

The winners

Neural Labs Africa and Me-Nasoma were named the two winners of the AI Academy Kenya Pitchathon 2026. Each takes US$5,000 in equity-free funding and Meta advertising credits, and each will pitch at Meta's AI Summit in Istanbul on 24–25 November 2026, themed "AI for All", alongside winners from Meta's other markets across Africa, the Middle East and Türkiye.

Two very different problems, one pattern: a Kenyan team, a tool people already use, and open models doing real work in the product.

Thank you to all ten for the work behind the pitches, and to the partners who made the day: Meta, the State Department for ICT and the Digital Economy, Kaara Works, Strathmore University's @iLabAfrica and The Social House.

What happens next

The second part of the Academy starts on Tuesday 3 November: an eight-week online Masterclass for 300 Kenyan students, startup founders, career professionals and developers, every Tuesday evening. Applications closed on 15 October; shortlisted candidates hear from us by 23 October and accepted candidates on or before 30 October. And in late November, two Kenyan startups take the stage in Istanbul.

If you pitched, applied, judged, spoke or simply turned up on Friday: thank you. Kenya's AI builders showed up, and so did the problems worth solving.