Notes from the insurtech panel at Africa Tech Summit Nairobi — separating the trends that move penetration from the ones that just look good on stage.
20 August 2026
At Africa Tech Summit Nairobi I sat on a panel titled "Insurtech — What's Trending?" alongside Jihan Abass, who founded Lami, and Elias Omondi. It is a deceptively hard question. "What's trending" invites you to list the shiny things — the new AI models, the new products, the new funding rounds. But in a market where insurance penetration is still stuck in the low single digits, the more useful question is narrower: which of these trends actually moves the number, and which just photograph well on a conference stage?
Here is how I answered it, and how I'd answer it again.
Embedded is the trend that survives contact with reality
If I had to name one trend that is genuinely changing the shape of the market, it is embedded insurance — cover sold inside another transaction rather than as a standalone product.
This is not a fashion. It is a direct response to the hardest problem in African insurance, which is not underwriting and not pricing. It is distribution against a customer who has already decided not to buy. Asking someone to stop, weigh an abstract future loss, and part with money today is asking them to lose an argument they settled long ago. Attaching cover to a moment when they are already acting — taking a loan, buying a phone, boarding a ride, shipping goods — removes the standalone decision entirely.
At Octamile we built API-first precisely so that the insurance disappears into the platforms people already use. The trend worth betting on is not "insurtech" as a category of standalone apps. It is insurance becoming a feature of everything else.
AI in claims and underwriting — real, but pointed at the wrong target if you're not careful
The room wanted to talk about AI, and rightly. It is the most consequential thing happening in the stack. But there is a difference between AI that reduces your cost of doing business and AI that changes whether customers trust you.
The obvious use is efficiency: faster underwriting, automated document review, fraud signals, straight-through claims. Those matter — margins in this market are thin and manual claims handling is where trust goes to die. But the deeper prize is using automation to pay claims before the customer expects it. In a low-trust category, the story that travels through a community is not your marketing. It is a neighbour saying they actually paid, and fast. AI that compresses a three-week claim into a same-day payout is not a cost play. It is a distribution play wearing an operations costume.
The hype filter: distribution and trust beat product breadth
The trap on any "what's trending" panel is to conflate activity with progress. There is a lot of activity — more products, more launches, more decks. Very little of it addresses the two things that actually gate this market: getting the product in front of people at the right moment, and being believed when a claim comes.
So my filter is simple. When someone shows me a new insurtech trend, I ask: does this make cover easier to reach, or does it make a payout more believable? If it does neither, it is a feature, not a trend. Most of what gets applause fails that test. Embedded distribution and fast, legible claims pass it — which is why they keep showing up as the through-line, panel after panel, year after year.
Where this goes
A market where most people are genuinely uncovered is not a small pool to fight over. It is a category to build. The insurtechs that matter over the next decade won't be the ones with the widest product catalogue or the flashiest model. They'll be the ones that made insurance effortless enough to be bought without a decision, and fast enough to be trusted after a loss.
That was my honest answer on the panel, and it is the bet Octamile is built on.
