Why we exist.
If you are 21 and live in Munich, San Francisco or Zhangzhou, and you have built a real solution to a real problem, the next step is simple. You apply to one of the many firms that back ambitious founders at pre-seed (sometimes when they only have an idea!). Then you get access to hundreds of thousands, sometimes millions, of dollars, euros or yuan which get you started and a vast network of like-minded founders to bounce ideas off of and learn from.
Now flip the story. Same age, same prototype, same early traction. But you were born in Stellenbosch or Nairobi. Here, to be taken seriously you almost need to be a four-time founder with twenty years in the industry. Fresh out of university with a prototype and a user or two? If you were born in Africa, unlucky for you. Same market, same risk profile as a founder in Sweden or Silicon Valley, but very few people willing to write the cheque. Not because you are weaker, but because too many investors still file a whole continent under "risky" and "no infrastructure".
The numbers tell a different story. Africa is home to 19% of the world's people and 28% of its under-25s. By 2100 it will be home to 46% of them, nearly half of all the young people on Earth [1]. The median African is about 19 years old [2]. Ten of the world's twenty fastest-growing economies this year are African [3].
And yet, in 2025, African start-ups raised about $2.2bn in equity. That is 0.4% of the roughly $470bn raised worldwide, and about the same as Sweden on its own [4].
So we have the youngest, fastest-growing continent on the planet, full of problems that need real solutions, and almost no one willing to write the first cheque. Too much of the early money that does arrive rewards whoever tells the saddest story rather than whoever builds the hardest.
Our founders do not want charity. They want people who back them and activate them. When the only way to get funded is to make up a story about how tough your life is, the best builders leave. They move overseas, or they take the graduate job at the big bank that headhunts on campus and climb the corporate ladder for the rest of their lives. The ecosystem never gets its snowball effect, and the best companies get built somewhere else.
Now give those founders the most powerful technology of their lifetime. They do not just catch up. They get supercharged.
People see the gap between Africa and San Francisco as a weakness. We see room to do what Africa is known for: leapfrogging.
Look at banking. Africa has about one fixed telephone line for every hundred people [5], and for most Kenyans a bank branch was never coming. Then in 2007 Safaricom launched M-Pesa, and people started sending money with a basic phone and a text message [6]. In 2006, 26.7% of Kenyan adults had access to formal financial services. By 2024 it was 84.8% [7]. Today about 41 million people use M-Pesa every month in Kenya alone [8]. Kenya never built the West's banking system. It went straight past it.
AI is the next leapfrog, and it is a much bigger one. A doctor, a lawyer, a teacher or an agronomist used to take years to train and a building to work from. Now a young founder can put that knowledge into a phone. The places with the fewest doctors, lawyers and teachers have the most to gain. That is us.
Oakvale exists to back the young Africans building with it, early, when conventional capital still will not.
"As we develop as a community, it is crucial for us to make sure that we are not just consumers of the latest machine learning techniques that have been developed elsewhere."
What we back.
We back serious solutions to real problems with immense potential, built by the highest-potential founders we can find. It is all about the founders.
We have three rules. A company has to meet all three or we pass.
- African founders, building from Africa. Win here first. Then take it global.
- A young founding team. The average age of the founding team has to be 18 to 25. We hold this line even when people push us to widen it.
- AI-native, at pre-seed. AI is the product, not a feature.
Our test for AI-native is simple. Switch the AI off. If the product still works, it was never AI-native. We will not back a ChatGPT wrapper, and we will not back an ordinary business with an AI dashboard bolted on the side.
We look at any sector. The ones that excite us most are the unglamorous ones: agriculture, law, education, logistics, infrastructure. Silicon Valley is pouring millions into AI agents that order your coffee. We see not being first as an advantage: we get to watch the mistakes and spend our money on problems that matter. Here the hard problems are still wide open. That is where the local data lives, where the workflows are messy, and where a good company becomes very hard to copy.
Why 18 to 25.
We do not back young founders despite their age. We back them because of it.
Africa is full of problems that need new solutions. We believe those solutions will come from the youth. Not one day. Now.
Half of Africa is younger than twenty [2]. The people who will live longest with the continent's problems are the ones most impatient to fix them. They have not yet been taught what is supposedly impossible (leading to a case of beautiful optimism and naivety), so they go and build it.
They do not need permission, a twenty-year CV or a sad story. They need someone to believe in them early, and a first cheque that goes a very long way.
Plenty of people have laughed when we tell them our mandate. We hold it with immense belief. To us it is the most obvious opportunity of our generation.
We are young too. We know what it feels like when nobody takes you seriously. We are here to be the people who do.
So why 18 to 25? Because age is the bet.
A long runway. A 21-year-old has a whole career of building ahead of them. Back them now and you might be backing their first company, their third and their tenth. That relationship compounds for decades.
Fewer habits to unlearn. Young founders are still working out how they build. Give them capital, mentors and people who have done it before, and they move fast.
Someone has to go first. Right now a talented 21-year-old in Cape Town is told to take the corporate job and play it safe. We want building a company to be a real option, not a long shot.
"We are going against the current, trying to convince young, smart people: you can start a startup. It is a normal thing to do."
Henrik puts it simply. Young founders do not need to be held by the hand. They need a North Star, honest counsel and the room to grow into it.
"True innovation will come from the younger generation."
The moat is everything.
Anyone can build a working prototype in an afternoon now. So building is no longer the advantage. The advantage is whatever is hard to copy once you have built it. That is the first question we ask every founder: what is your moat?
The answers we like:
- local data that gets better the more people use the product
- being built into how a customer works every day
- real depth in a domain or its regulation
- distribution nobody else has
- customers who would find it painful to leave
Sometimes speed is the only lead a company has. That can work, but the speed has to be real and it has to last.
The hardest version of the question is the big-lab one. Models keep getting cheaper: the cost of running a model as good as GPT-3.5 fell more than 280-fold between November 2022 and October 2024 [10]. OpenAI, Anthropic and Google keep folding the obvious features into their own products. When that happens, is this company still standing? If the only answer is "we got here first", that is a head start, not a moat.
Building version one on a frontier model is fine. We expect it. The data flywheel, the company's own models and the real defensibility come after, once there is local data and real usage to build them on.
Where we look.
We are layer-agnostic. We do not start by asking where a company sits on the AI stack. We ask whether the AI is essential, whether it solves a real local problem, whether there is a moat, and whether this founder can move faster than anyone else. The layer follows from that.
We are sector and layer agnostic within AI. That covers applications, vertical tools and infrastructure: we follow the founder and the problem.
How we invest.
Patient on the arc. Impatient on proof. We will wait years for a company to matter. We will not wait a quarter to see a founder moving.
Henrik has a rule of seven: it takes about seven years to succeed, and there is about a seven percent chance you do. We invest knowing both numbers.
One milestone at a time. At the start we agree a single, company-specific proof point with the founder, usually a handful of paying pilots. Then we fund the leanest route to get there.
Disciplined cheques. More money does not buy product-market fit. A smaller cheque forces cheap mistakes early, and cheap mistakes are the best kind.
Flexible capital, lean teams. We often release capital in tranches against the milestones we agreed. We would rather see a founder hire hungry students and junior talent who own equity than burn the round on expensive senior hires before the business is proven.
We can do this because we are a single-LP family fund. There is no return clock forcing us out early. We can wait for a company to be ready.
We activate exceptional builders.
We do not want to be a name on a cap table. Backing someone this early means building the founder as much as the company.
We activate exceptional builders. The founders we back are already extremely capable builders. Our job is to equip them with the business sense, network and operating discipline to turn that into a company.
Our founders get Henrik. More than two decades of starting, running and selling companies [12], and a door that is always open. They also get our network of customers, talent, advisers and follow-on investors.
We stay close from day one. Clear milestones, help with positioning, go-to-market, hiring and getting ready for the next raise, and a shared space to work from.
A founding batch, not a portfolio. Our founders know each other, share mentors and build next to one another.
We back you with belief, capital and everything we've learned. In return, we ask for three things. Obsession with the problem, not just the company. Honesty with us, especially when things go wrong, because bad news early is something we can help with and bad news late usually isn't. And the ambition to build something that matters far beyond Africa, from Africa. Do that, and we'll be in your corner for the long run.
How we operate.
These are the things we hold ourselves and our founders to. A way of thinking, used with judgment, not a rulebook.
Know the one thing that matters. The best founders are clear on the single goal and ignore the thousand distractions around it.
Build in public. Nobody backs what nobody can see. Hiding the work, waiting for it to be perfect and being scared to fail cost far more than they protect. We think fear of failure holds back more African founders than a lack of capital does.
Failure is part of building. We have no time for the sob-story version of African entrepreneurship. We want pride, ownership and momentum. The question we keep asking is: what did you build, what broke, and what did you do next?
Foundation before spotlight. A few strong companies earn us the right to get loud. Good investments bring the next ones.
Start local, think global. Africa is not one market. South Africa alone has twelve official languages [13]. So we start in South Africa, put down roots and build deep networks, then take it outward.
The more we use AI, the more human we need to be. Taste, judgment and relationships do not get automated. That is where we compete.
Keep the bar high everywhere. In who we back, who we hire, what we publish and how we host. Young does not mean lower standards.
The fund.
Oakvale Invest is a pre-seed fund based in Cape Town, with R200m committed to African AI founders.
Today that capital comes from a single family LP. That is what lets us be properly patient. Our first cheques are roughly R1m to R3m, in preference equity, often staged in tranches against agreed milestones. We focus on a few strong companies before we go wide.
Everyone wants to invest in Africa. Very few know how, because you have to be on the ground. So next, we are raising a separate seed fund, with our own capital as its anchor, for international family offices and investors who want real exposure to African AI and to the highest-potential founders building it. It also gives our best companies follow-on capital at seed. Same mission. Next chapter.
Sources for every figure.
- Pew Research Center, "5 facts about Africa's population growth", 19 May 2026 (UN World Population Prospects medium variant). https://www.pewresearch.org/short-reads/2026/05/19/5-facts-about-africas-population-growth/
- UN World Population Prospects 2024, via Worldometer and Statista: Africa median age 19.2 (2024). https://www.worldometers.info/world-population/africa-population/ and https://www.statista.com/statistics/1224179/population-growth-in-africa/
- IMF World Economic Outlook, October 2026, real GDP growth projections for 2026 (DataMapper NGDP_RPCH). Ten of the top twenty: South Sudan, Ethiopia, Rwanda, Guinea, Sudan, Niger, Benin, Uganda, Tanzania, Côte d'Ivoire. https://www.imf.org/external/datamapper/NGDP_RPCH@WEO
- Africa: The Big Deal, "Small but Mighty", 24 February 2026 (Africa data: ATBD; global and Sweden: CB Insights). https://thebigdeal.substack.com/p/vsworld25
- ITU, Measuring digital development: Facts and Figures 2024, fixed-telephone subscriptions per 100 inhabitants by region. https://www.itu.int/itu-d/reports/statistics/wp-content/uploads/sites/5/2024/11/2402588_1e_Measuring-digital-development-Facts-and-Figures-2024_v4.pdf
- Safaricom, M-PESA (launched in Kenya in March 2007). https://www.safaricom.co.ke/main-mpesa/about-m-pesa/m-pesa-journey
- Central Bank of Kenya, KNBS and FSD Kenya, 2024 FinAccess Household Survey. https://www.centralbank.go.ke/wp-content/uploads/2024/12/2024-FINACCESS-HOUSEHOLD-SURVEY-MAIN-REPORT.pdf
- Safaricom FY26 results booklet, 7 May 2026: 40.99m one-month active M-PESA customers in Kenya. https://www.safaricom.co.ke/images/Downloads/FY26-Earnings-Booklet-2026-7th-May-2026.pdf
- Benjamin Rosman, "To our Success", Deep Learning Indaba, signed November 2023 (published 20 January 2024). https://deeplearningindaba.com/blog/2024/01/to-our-success/
- Stanford HAI, AI Index Report 2025: inference cost at GPT-3.5 level fell from $20 to $0.07 per million tokens, Nov 2022 to Oct 2024. https://hai.stanford.edu/assets/files/hai_ai_index_report_2025.pdf
- Chelsea Adrian, Varsity Newspaper, "A R200 Million Bet on South Africa's Next Generation of Tech Startups", 16 April 2026. https://varsitynewspaper.substack.com/p/a-r200-million-bet-on-south-africas
- Wikipedia (sv), Henrik Persson Ekdahl: founded Park Trade Europe in 2003. Finansavisen, 20 December 2011: then head of Betsson's Malta operations, at age 31. https://sv.wikipedia.org/wiki/Henrik_Persson_Ekdahl and https://www.finansavisen.no/nyheter/livsstil/2011/12/vi-er-ingen-bad-guys
- South African Government, Constitution Eighteenth Amendment Act 2023: South African Sign Language becomes the 12th official language (signed 19 July 2023). https://www.gov.za/news/media-statements/minister-zizi-kodwa-south-african-sign-language-being-recognised-official
Follow our journey on LinkedIn, Instagram, YouTube and Substack.
