Most African venture capital firms make fewer than ten investments each year. But arriving at that number requires reading through hundreds of pitch decks, replying to a good number of founders, and multiple conversations before a final decision is made.
The problem is that your average Africa-focused VC firm isn’t staffed by a lot of people. Even more importantly, reviewing and analysing pitches is just one aspect of the work a VC has to get through. There are portfolio companies to support, investors to report to, founders to source and markets to understand.
For Oui Capital, an early-stage investor that has stayed small since its launch, these considerations were at the heart of its decision to not just embed artificial intelligence into its investment workflow but to build an internal tool that leverages data from more than six years of investing history.
“AI is how a tiny team punches above its weight: more ground covered, faster, without adding headcount,” Francesco Andreoli, co-founder and venture partner at Oui Capital, shared with Condia.
Why build an AI investor assistant?
The artificial intelligence landscape today is littered with tools for practically anything. Need to take and summarise notes in a meeting, manage your finances or automate a repetitive task? There is probably an AI tool for it.
For investors, and specifically venture capital investors, the options are fewer and often designed for narrow use cases.
Before deciding to build an internal tool, Andreoli, who has an engineering background, shopped around for products but often found tools that served specific needs without addressing the investment process as a whole.
“Today, you have very advanced CRMs, but you don’t have a unified platform that is specifically built for venture capitalists.”
The decision to incorporate AI into the investment process was not simply about following a fad. A major consideration for the team was encoding its experience from the last seven years.
Investment decisions are shaped by years of accumulated judgement: which markets the firm understands, which types of founders it backs, what signals it considers meaningful and, just as importantly, which warning signs it has learned to recognise.
Much of that knowledge traditionally lives inside the heads of partners. Oui Capital’s bet is that some of it can be made legible to a machine, and the result is Oui Capital OS.
What Oui Capital OS does
At least twice in the first ten minutes of our conversation, Andreoli reiterates that Oui Capital OS does not determine whether the firm passes on a founder or not.
Instead, it is designed to augment the investment team, rather than replace it. Its work begins long before a startup or founder first interacts with the firm. It crawls the internet, searching social media, product marketplaces, media publications and other sources to surface startups that might fit the firm’s investment strategy.
The next stage is pre-screening. Oui Capital receives tens of pitches each week, Andreoli says. A year ago, analysts had to read through each deck to identify which startups the firm wanted to investigate further.
Today, the AI assistant can screen companies for factors such as the markets they operate in and their traction before passing relevant opportunities to the investment team. That may sound like a relatively mundane application of AI. In venture capital, however, the ability to process more opportunities can have an important consequence.
The traditional constraint on a small VC is not necessarily a shortage of opportunities. It is attention.
There are more founders than a small investment team can meaningfully evaluate. AI therefore does not need to discover a hidden unicorn to create value. It can simply allow investors to spend more of their scarce human attention on the companies that deserve it.
The biggest gains, however, appear to be coming later in the investment process. Oui Capital uses AI extensively during due diligence, gathering information from external sources and working through material shared by startups. Tasks that previously took weeks or months can now be completed in a matter of days.
“We’re cutting down the data room process to days. Before it would have taken weeks,” Andreoli shared.
The firm is also seeing value before formal due diligence begins. Instead of relying on multiple calls and interactions with founders to uncover information that could influence an investment decision, the team can identify some of those signals much earlier.
Together, these changes have cut the average time from the first call with a startup to investment by roughly half.
The human advantage
If a machine can read pitch decks, analyse markets, conduct research and process a data room, what exactly is left for the investor to do?
Andreoli’s answer is, in some ways, the opposite of the techno-optimist view that AI will eventually automate the entire profession.
Individual experience may become more valuable, not less, he shared. This is particularly true in markets where reliable data remains scarce. African startups often operate in markets where historical datasets are limited and information fragmented.
AI can find and organise information, but it is less capable of knowing which piece of incomplete information matters.
That distinction is important for venture capital because investing is not simply an exercise in analysing what has already happened. Investors are making bets on what might happen.
A spreadsheet can tell an investor that a company is growing quickly. It cannot, on its own, tell them whether the founder can navigate a difficult market or period, whether customers genuinely love the product or whether a particular regulatory change will transform the business.
Those judgements are precisely where experience, relationships and intuition still matter. There is another potential advantage.
By encoding its historical investment decisions into a system, Oui Capital is effectively attempting to turn institutional memory into infrastructure.
That could prove useful as the firm grows. People leave firms, partners change responsibilities and institutional knowledge can become diluted. An internal AI system gives the firm another way of preserving how it thinks.
There is a risk, of course. If an AI system learns primarily from a firm’s historical decisions, it can also learn the firm’s historical biases. A machine trained to recognise the characteristics of companies previously backed by a VC could become very good at finding companies that look like the past but fail at identifying companies that represent something genuinely new.
For a profession built around finding outliers, that could be a serious limitation. Which is why the human investor remains necessary, even as the volume of work they personally do falls.
The one-person VC firm
The venture capital industry has spent years talking about the possibility of the one-person unicorn. AI raises a similar possibility for venture capital.
Could the next generation of funds manage billions of dollars with dramatically smaller investment teams? Perhaps. But the more immediate change may be less dramatic and more consequential.
A small African VC does not need AI to replace its partners. It needs AI to give those partners more hours in the day, more visibility into the companies being built around them and a better way of applying what they have learned.
That could change the economics of investing in markets such as Africa, where funds are often small, opportunities are geographically dispersed and the cost of doing deep diligence can be high.
The competitive advantage may therefore not belong to the firm with the most sophisticated AI. It may belong to the firm that best combines AI with a deep understanding of its market.
For Oui Capital, that experiment is already underway.
Last updated: September 15, 2026
