Advertisement banner image

“Solve boring problems,” African AI leaders say at Builders Summit

For builders in the room keen on learning about areas to focus, the advice from the panel of experts was clear: solve repetitive problems.
6 minute read
“Solve boring problems,” African AI leaders say at Builders Summit

Ask anyone in global tech what’s hot right now, and you’ll get a predictable answer. Microsoft, Meta, Google, OpenAI, all racing to dominate AI. Whether it’s through new chips, enterprise copilots, or custom LLMs, the signal is loud: AI isn’t just a feature anymore. It’s the strategy.

And yet, when you zoom into Africa’s startup ecosystem, that signal can feel distant.

“It’s not about copying trends,” said Hassan Luongo, partner at Resilience17, an AI-focused venture studio. “But the sheer weight of activity in AI, globally, makes it more than a passing cycle. It’s a foundational shift.”

He pointed out that AI-first startups now dominate entire Y Combinator batches. OpenAI is moving into everything from cloud computing to devices. Even global VCs are rewriting their investment theses to prioritise AI integration.

The question isn’t whether AI matters. It’s how we make it matter here.

Sponsored Ad Sponsored

That was one of the first things said during a panel at the 2025 Builders Summit by Founders Connect, and it landed with a quiet weight. The speakers were bullish on its possibilities. But the message was clear: cut through the noise. If African founders are going to build meaningful, defensible startups in this AI moment, they need to stay focused on fundamentals.

Moderated under the title “Mapping the AI Ecosystem: Adoption Patterns Among African Startups”, by Daniel Adeyemi, Condia’s Editor-in-Chief, the panel brought together voices with hands-on experience: Ebuka Obi, CEO at Autogon; Kehinde Olateru, co-founder at Zero Complex AI; and Hassan Luongo, partner at Resilience17.

The 30-minute conversation before hundreds of tech builders in the hall wasn’t about buzzwords, but creating a better understanding of what’s happening in the African AI space.

Building from scratch

A recurring point was how AI adoption in African startups is, at least for now, largely dependent on plugging into existing models and infrastructure, not building Large Language models (LLMs) from scratch. The most obvious reason for this is due to the expensive nature of training LLMs. General-purpose language models like GPT-3, Claude, Mistral, or LLaMA can cost anything from $500,000 to as much $100 million. GPT-4 is believed to have cost over $100 million to train due to massive scale, safety training, and reinforcement learning steps.

Although Africa is home to more than 2,400 AI companies, according to an Afrilabs report, African AI startups raised just $4 million across five deals in Q2 2024, per data from CB Insights. That’s a stark contrast to the $23.2 billion raised by their global counterparts during the same period.

“Training your own foundation model? That’s not happening here anytime soon,” said Olateru. “But there’s power in layering AI tools on top of real-world problems.”

Obi agreed, adding that local startups are integrating with APIs from OpenAI, Google, and others to solve sector-specific pain points — in fintech, healthtech, edtech, and beyond.

Take fraud detection, for example. One financial institution uses AI to scan millions of transactions in real time, flagging anomalies faster than any human could. It’s not revolutionary tech in itself, but in a local context where even a single fraudulent transaction can severely damage trust and operations, the impact is massive.

Read also: ZeroComplex AI is simplifying AI integrations for African businesses

Solve boring problems

For builders in the room keen on learning about areas to focus, the advice from the panel of experts was clear: solve repetitive problems. Repetitive, everyday tasks that waste time and money — the kind of inefficiencies that compound in under-resourced environments.

Obi shared how one Nigerian bank transitioned from a full human-staffed call centre to AI-powered voice assistants, drastically reducing costs while handling higher call volumes.

Olateru built on that, arguing that the greatest AI opportunities on the continent don’t lie in headline-grabbing moonshots but in systems that make daily work easier.

“Every founder here could probably name a repetitive task in their sector that’s still being done manually,” Olateru said. “That’s your AI use case.”

During the session, the audience was introduced to a handful of startups from Resilience17’s AI accelerator cohort. Each one embodied what the panel had been arguing for: not flashy tech for tech’s sake, but purpose-driven tools rooted in local context.

There was a startup building AI tools for Nigerian teachers, helping with grading, lesson planning, and general classroom management. Teachers in overcrowded public schools, often underpaid and under-resourced, could use these tools to reclaim their time and attention.

Another company had launched a mental health chatbot that speaks multiple Nigerian languages via WhatsApp, combining culturally relevant conversation with access to live therapists when needed.

A third, Tyms, was focused on small business accounting, letting entrepreneurs query their own financial data in natural language: “How much do my customers owe me this month?” “What were my top three expenses last quarter?”

And then there was the legal tech play, using AI to draft standard contracts for startups and SMEs at a fraction of the usual cost, without sacrificing legal soundness.

“They’re not building generalist tools,” said Hassan. “They’re solving specific problems with tailored intelligence. That’s the difference.”

Building an AI startup is still about the basics

Despite all the talk of AI models, the panellists repeatedly came back to a central idea: building an AI-powered startup is still building a startup. The basics haven’t changed.

“What does your onboarding look like?” Hassan asked. “Is it clear? Frictionless? Does your user understand what you’re doing in 15 seconds or less?”

Resilience17’s focus, he explained, wasn’t just on introducing founders to AI techniques. It was helping them sharpen their value proposition, fix product flows, and think deeply about user acquisition and retention.

That’s what most founders actually need — not a crash course in deep learning, but support in applying AI meaningfully while still obsessing over product quality.

The summit was filled with optimism, but it wasn’t naive. There were no sweeping declarations that AI would “fix” Africa. 

Instead, what emerged was a picture of thoughtful ambition. A generation of builders who understand that the future is being reshaped by AI, but who also know that execution, clarity, and context still matter more than buzz.

“The most dangerous thing founders can do right now is chase the hype without doing the work,” said Hassan.