Askya Investment Partners ha aperto le candidature per un programma di sei settimane destinato alle startup africane di AI, con un impegno minimo di $200,000 per almeno un partecipante. Dieci aziende prenderanno parte al programma e nessuna di esse dovrà pagare in contanti né cedere equity per partecipare.

Fondata nel 2024 da Babacar Seck, la proposta della società si basa in larga misura sul suo percorso professionale. Seck ha lavorato per anni in AXA come advisor del chairman e chief executive prima di diventare membro fondatore del programma di venture capital da $300 million di Proparco e, in seguito, chief executive di Digital Africa. In questi ruoli ha sostenuto quasi 20 startup — tra cui Moniepoint, Jumia, GoMyCode e Complete Farmer — ed è stato parte della prima generazione di gestori istituzionali di fondi di venture capital nel continente. Secondo la società, quegli investimenti hanno generato oltre $120 million di profitti per gli investitori, incluso un listing sul New York Stock Exchange. Ha inoltre presieduto un gruppo di lavoro sugli investimenti del consiglio AI di Smart Africa, contribuendo a definire una strategia continentale per gli investimenti in AI.

Seck è molto preciso nel descrivere il programma. Dice che non si tratta di un accelerator, perché gli accelerator sono costruiti attorno alla raccolta fondi. Insegnano ai founder a fare pitching, a rifinire il deck e a concludere con un demo day che di fatto è una giornata di presentazione agli investitori, trattando la raccolta come il risultato finale.

Il programma di Askya parte da una diagnosi diversa. Molte aziende africane trovano una domanda reale per i propri prodotti ma faticano ancora a scalare perché le fondamenta sottostanti sono deboli — governance, assunzioni e tecnologia. Secondo Seck, questo raramente dipende dall’intelligenza o dall’integrità dei founder. È piuttosto una questione di esposizione. Il programma affiancherà quindi i founder a operatori e builder che potranno fare coaching, insieme a masterclass su go-to-market, distribuzione, pricing, governance e tecnologia.

Il secondo problema che Askya vuole affrontare è la disconnessione. Seck descrive un ecosistema in cui le varie componenti non comunicano tra loro. I data center inseguono i clienti mentre tutti sostengono che l’Africa abbia bisogno di più data center. I provider globali di cloud segnalano in modo discreto che alcuni dei loro maggiori clienti africani sono startup, con alcune scale-up che spendono oltre $30 million all’anno. Aziende e governi dicono di voler lavorare con startup africane ma non sanno quali siano valide. Le università formano ingegneri AI che poi trovano lavoro all’estero, perché sul continente troppe poche aziende operano alla scala necessaria per trattenerli.

Askya porterà nel programma telco, banche e aziende tecnologiche per presentare ai founder problemi concreti. Sta inoltre collaborando con Deep Learning Indaba per collegare le aziende partecipanti alla più ampia comunità africana dell’AI.

Su ciò che Askya cerca davvero, Seck è altrettanto preciso. Le aziende devono essere fondate in Africa e basate nel continente, con founder a tempo pieno, un prodotto funzionante e i primi clienti. Non possono essere in fase di idea, ma Askya non richiede nemmeno milioni di ricavi. La società è aperta ad aziende fino alla fase pre-Series A.

Oltre a questi requisiti di base, Seck afferma che ci sono tre test. L’azienda sta risolvendo un problema reale oppure sta costruendo una soluzione in cerca di un problema? Il prodotto è significativamente migliore delle alternative, considerando ciò che i clienti africani valorizzano davvero? E il team ha ciò che serve per costruirlo?

Il quarto criterio — quello a cui Seck torna più spesso — è l’impegno. Askya non cerca founder alla ricerca di un rapido investimento e di un exit. Internamente, la società definisce il tipo di aziende che vuole costruire come “tech Dangotes”: imprese generazionali capaci di durare decenni.

Come osserva Seck, costruire una grande azienda non richiede cinque anni. Richiede sette, 10 o 15.

L’intervista qui sotto è stata leggermente modificata per lunghezza e chiarezza.

What does your ideal African AI startup look like?

Babacar Seck: When we talk about why AI is interesting to us, the fundamental reason is that AI is a productivity enhancement tool. That is really the core of it. Everything we see and like derives from that, because if you look at the impact of AI in the world today, where it is impactful, it is driving up productivity and effectiveness.

We think about it as a tool that can help us solve our problems in energy, healthcare, industrialisation, economic inclusion, and other sectors.

If we zoom in on where we have seen exciting startups, I would separate them into two blocks. Internally we call the first one the AI stack — the different layers of the AI value chain. One very exciting area is voice AI in African languages. You have probably seen many startups operating in that space. They have not fully cracked it yet, but think about it: today in Africa, most people do not speak English, French, or Portuguese at home. They speak their mother tongue, the language they are more comfortable in. The data in South Africa shows 80% do not speak English at home, which surprised many people, including us.

The minute you have an application — and several are being developed — that lets companies interact with their customers in the customer's own language, you are bridging a divide. Not only in company-to-customer interactions, but also in government-to-citizen ones. We are starting to see this with NGOs using WhatsApp voice messages in African languages to reach people who are often not literate and who live in rural areas where English or French education is less developed. It is a tool with the opportunity to bridge that divide.

The other block we call AI for the real economy. This is the pure productivity play. Everywhere you have structured data at scale, you can build innovative small AI models that enhance performance. This is very exciting in the African context, because we are digitising in the age of AI. We are creating structured pools of data at the same time as we are applying AI to them. It creates a leapfrog effect, because you are building an economic moat much faster as you scale.

Think about the energy space. We see B2B companies in Nigeria that help you optimise your energy installation. Because of the weakness of the grid, people overcompensate, especially for critical industrial operations that cannot go down — telecom towers, manufacturing, and so on. You oversize on solar power, batteries, and generators, and you do not optimise, so you spend a lot more on energy consumption. With AI you can optimise that, using not just data from the platform but also external data — weather, the cost of diesel today, the cost of the grid, and the temperature — and work out how to optimise the mix. You can right-size and also optimise day-to-day use.

We see tangible impact. This is not talk. We see 20% to 30% cost reductions on energy consumption for businesses using such solutions, which is huge, and it has even more potential as the AI gets better.

Another example emerging as a trend is health insurance. Health insurance is not a big thing in Africa today. Even in Nigeria, one of the continent's biggest economies, maybe 5% of adults are insured. One of the reasons is that it is very expensive. If you go deeper and ask why, it is because there is so much fraud. I come from the insurance sector, so I know this firsthand. No insurer can tell you exactly how much fraud they have, but it is generally estimated to be at least half of claims. Many insurers either lose money or do not make big margins on health insurance because of fraud, and that is also why they do not expand aggressively. In Nigeria it is mostly companies insuring their employees. People do not go out and buy health insurance for themselves because it is not affordable.

What companies do with AI in that space is leverage image recognition, which has become much more advanced. They can authenticate documents and proof of a medical act with more accuracy and make immediate payouts, which you would not do if you had high suspicion of fraud. You may have experienced going to a hospital with insurance and still being asked to pay, even within the network, because the insurer is going to take a year to pay them. This changes the whole working capital flow in that entire sector, and we see companies doing it at a growing scale.

Only one startup gets a cheque. What do the other nine get?

Babacar Seck: You noticed we did not call this an accelerator. It is an AI growth platform. Most accelerators are focused on fundraising. They teach you how to pitch, how to build a good deck and business plan. You get some coaching, and at the end you have what they call a demo day, which is not really a demo day — it is a pitching day. You pitch to investors and you raise, and the raise is the outcome of the programme.

Our programme is structured differently because we saw two problems in the ecosystem.

The first is that many companies that manage to get demand for their product — what we now call product-market fit — have weak fundamentals, and that stops them from growing into enduring companies. Whether it is governance, managing people, or technology that does not scale and breaks at some point. Most of the time those are not the result of founders who are unintelligent or dishonest. It is limited exposure to people who have already encountered those challenges.

We address that by centring the programme on builders and operators coaching entrepreneurs. The people coaching in the programme have built businesses themselves. It is a mix of coaching and masterclasses over the six-week programme, giving founders a strong foundation on those topics and on the technology side, guiding them on how to build companies that endure and scale. There is a big focus on go-to-market, distribution, sales, and pricing, with that core layer of governance, technology, and company building underneath. Every company that goes through the programme gets that, and between you and me, that is the highest value, because those are the things you cannot otherwise get. You cannot access those people unless you know them, outside a structured programme.

The other challenge we saw comes from being deeply involved in the AI ecosystem — through advocacy in the Financial Times and the Africa CEO Forum. I also chaired a Smart Africa AI council investment working group, where we devised an investment strategy for African AI at the continental level. One thing we noticed is that all the big issues we face, with the exception of regulation, come down to companies not scaling.

When you speak to infrastructure players and data centres, you realise most data centres are chasing customers, yet we say we do not have enough data centres in Africa. One of the gaps is that these operators are not plugged into demand. But when you speak to the global cloud players — and because we are on record I will not say which ones — they will tell you their biggest customers in Africa are actually startups, with some scale-ups spending upwards of $30 million a year on cloud.

You see the same disconnect with corporates and government. They say they would love to partner with African startups, but they do not know them. They do not know who is good. They see some news online, they engage, some of them have accelerators as a more structured way to engage, but it has not been translating.

And when you speak to universities, they tell you they cannot find jobs for the talent they train. We are training a growing number of AI engineers — at Carnegie Mellon in Kigali, at the African Institute for Mathematical Sciences, and at some really good Nigerian universities — and those people all find jobs abroad, because the platforms on the continent do not exist at the right scale to retain them.

All of this comes back to scaling the startups while connecting them to the ecosystem. That is why this platform is also about connecting to these nodes. We are partnering with Deep Learning Indaba, the largest community of AI professionals, researchers, and learners on the continent, which held its event in Lagos last week — we were there — and a few other organisations, so we can plug these startups into the ecosystem that can scale them.

It is the coaching, the plugging into the ecosystem, and the investment is the cherry on the cake. We are not saying we will only invest in one. We might invest in several, but our commitment at a minimum is one.

People are not coming for the money. We want to attract people who want to scale their businesses and give them solid foundations.

Does the $200,000 come from your fund? What are the terms?

Babacar Seck: It comes from our fund. The terms will be on a case-by-case basis, so it is fully flexible.

What is the selection process, and what are you looking for?

Babacar Seck: Let me describe the process, then tell you what we optimise for.

We have an online application form with a few questions that help us understand whether companies meet the high-level criteria. Those are: African-founded, based on the continent, solving problems here. Founders need to be full-time, need to have a product, and need to have first customers. It is not idea stage, but we are also not requiring millions of dollars in revenue. We are flexible on the maximum stage, up until pre-Series A. The minimum is showing some level of execution and commitment to your business.

In terms of what we optimise for, the first is driving productivity on the continent. Are they solving a real customer problem? Very often, when you dig in, companies are building a solution but not solving a problem. At the core is focusing on pain points that are significant and at scale. That tells us there is potential demand for what they are building.

Second, is it innovative? Is this meaningfully better than the alternatives people have, given customer preferences? In Africa, some telcos discovered it is better to optimise for cost rather than quality, because people are price sensitive. That is not the obvious thing you would do in other regions, but in Africa that is what matters. If you can vastly optimise for what customers seek most, you have something. In the Nigerian payments space, what people needed was access, distribution, and reliability, more than complex features. That is what merchants wanted, and the players who provided it now dominate the market.

Third, can they build it? Does the founding team have the skills required? If you are building an AI solution and nobody on the team has experience building it or any competitive advantage in doing so, that leads us to more questions.

One very important thing for us is commitment. Are you committed to solving this? We want demonstration of that, not just words. How have you been trying to solve this problem before you even built this business? Why do you care? We are not trying to do a quick invest and flip. I think it was TechCabal that ran an interview I gave, and you titled it "the exit is not the goal," and then I got calls from my investors — but we stand by it. The goal is building great, generational, enduring businesses. Internally we call them tech Dangotes. That is what we are looking for.

Nobody needs to teach you that it does not take five years to build a great business. It takes seven, ten, or fifteen at best. If you are not committed to doing this to your maximum over the next decade or more, you are likely going to give up. The level of the team's commitment to the problem is the most important thing.

Your deck mentions corporates. Why involve them, and which ones are involved?

Babacar Seck: We are finalising the partnerships, so we cannot announce names yet. The profile is telcos, banks, and tech corporations.

It is important to have corporates involved because they drive the demand in our market and they are extremely hard to read. We have seen startups engage with a telco for a year and a half, and nothing moves because of not having the right person on the other side or corporate complexity. Having them in the programme helps us create that route to market much faster. And the corporates are the ones with the problems — you want to be close to the problem for the startups to be able to innovate.

There will be workshop sessions run by corporates where they present their challenges and priorities, and startups get the opportunity to engage and propose solutions. Corporates can also be distributors, especially telcos, because of their customer base, and banks, because of regulatory factors. Both are very good partners to startups. It is about being close to the problem, and it is about distribution.

What talent do you have on the continent coaching these founders?

Babacar Seck: It is a mix of founders and operators.

For founders, we have people who have built and run businesses past $50 million in revenue coming to share their lessons learned — from success but also from failure — across multiple sectors. We have them from both Francophone and Anglophone markets to highlight the differences between those markets.

On the operator side, we have people like the head of AI for one of the hyperscalers in Africa and the founding CEO of Mansard, plus a few more from Senegal, Guinea, Kenya, and outside Africa.

In AI specifically, we looked for people working at big companies who have built products there and also built and scaled startups. We have people from NVIDIA and Google involved.

There are three categories really: the AI builders who have been founders and also worked in big tech, the founders I mentioned, and the operators. Among the operators we also have someone who heads technology for Egon Zehnder, the biggest leadership consulting firm, which helps companies hire CEOs and CTOs globally.

I asked some of them what a workshop like the one they are giving normally costs. It is typically around $50,000. Here, founders are getting it for free.

Why do you think this can be run in six weeks?

Babacar Seck: The most important things with the problem we are trying to solve are design and iteration.

The design is having the core principles in place: you are learning from builders, you are getting access to partners and customers, and it is not a distraction from your core business. Those are the three most important things. The rest is iteration.

We are starting with six weeks. This is something we are going to scale. It is not a one-off. It is a real commitment we are making. As I mentioned, we think AI is going to become one of the most important drivers of economic development in the world. It will determine the difference between who is rich and who is poor and which country grows and which does not.

This challenge is a today problem, even a yesterday problem. Africa has not committed to scaling its AI ecosystem the way it should. We have been advocating that Africa should be a producer of technology, not just an importer of these solutions; otherwise, our productivity gap with other economies will keep widening. We have been advocating for this, but at this point it is our responsibility to build it. This is version one. This is the start. Even in its current form, we think it can have a huge impact on the companies that go through it. We are putting our money where our mouth is, backing founders and bringing what they need to scale these businesses to solve Africa's needs — bringing painkillers, and I would argue antivenom, into the market against all these risks and needs people have.