Building the Intelligence Behind the Flights That Keep Africa Connected.
How Prometheus X Fellow Leroy Mwasaru is using AI to make regional aviation safer, smarter, and more reliable.
Across Africa, regional aviation is essential infrastructure. But the smaller operators connecting cities, businesses and communities often lack access to the sophisticated maintenance intelligence available to the world’s largest airlines. Leroy is building AirVara to close that gap.
AirVara is developing aircraft-maintenance intelligence for regional, charter and turboprop fleets, helping operators spot engine-performance problems earlier, manage maintenance and compliance, and keep aircraft flying safely and reliably.
Start with the problem. What is happening behind the scenes when a regional aircraft is unexpectedly grounded, and why is that such a consequential problem?
Everything stops at once. Engineers chase the fault, operations re-shuffle the schedule, and someone hunts for parts. Currently, the cost of keeping an aircraft grounded are unforgiving at $10,000 to $150,000 an hour, depending on the aircraft. For an operator with eight aircraft, one grounding takes out more than 10% of its capacity. From the planes we’ve monitored, most of these aren’t really sudden. The warning signs were in the engine data weeks earlier, scattered across spreadsheets and logbooks.
What did you see in the aviation industry that made you think, “there should be a much better way to do this”?
It stemmed from seeing press releases about delays and reduced capacity. And it got me asking my aviation professional network what was behind the issues highlighted. I also benchmarked a few of these operations in-person. The answer kept coming back: legacy maintenance platforms are expensive and built for fleets of hundreds, so smaller operators make do with spreadsheets. I wanted to give them tools that offered more at a fraction of the cost.
So what exactly is AirVara building? Explain it to someone who knows nothing about aircraft maintenance.
An early-warning system for engines – both turboprop and jet engines. Turboprop engines like the Pratt & Whitney PT6A record temperature and torque on every flight. A slow rise over a few hundred flights is easy to miss and often the first sign of trouble. AirVara flags that drift, shows the readings behind it, and points to the right page of the manual. It also tracks every inspection, airworthiness directive, and open defect. The engineer makes the call, and the executives plan a financial roadmap.
AirVara seems to be concentrating on regional, charter and turboprop operators rather than the largest airlines. Why start there?
They’re underserved, they matter, and they feel the problem most. Also, smaller operators serve as last-mile bridges for organizations like WHO, WFP, and others. Enterprise maintenance software is built for fleets of hundreds. In Africa, turboprops connect the secondary cities, mines, and humanitarian routes the big carriers don’t fly. The thesis for this is: An airline with 300 aircraft can absorb a grounding. An operator with eight can’t.
Where does AI actually matter in the product? What can AirVara detect or help an operator understand that would otherwise be difficult, slow or easy to miss?
AI matters in two places. First, catching slow drift across hundreds of flights that nobody would spot one flight at a time. Second, turning a flag into an action: AirVara pulls the relevant procedure from manuals that run to thousands of pages and cites the source. Every flag traces back to the readings behind it. Certified engineers make the airworthiness decision. AirVara’s AI shortens the time it takes to get there.
What have you learned from putting the technology in front of real operators? Was there anything customers taught you that materially changed what you were building?
There’s a lot of learning. Early users are telling us they make decisions faster with engine trends, compliance status, and manual references in one place. The biggest lesson is that the technology is the easy part, and the moving parts are more challenging and important to integrate into. AirVara has to earn a place in an engineer’s working day.
Your own path into aviation is unusual. You began building climate technology as a teenager, later worked in software and education technology, and are now building in aviation. What connects those chapters for you?
At school in Kenya, I co-built a system that turned human waste into biogas, and later founded Greenpact. Then came software: internal tools and games at a SpaceX lab school. The pattern is the same each time: find the fragile part of a system everyone relies on, and make it reliable.
Africa is often discussed as a future aviation-growth market. From where you sit, what does the rest of the world misunderstand about the infrastructure required to make that growth possible?
Africa is home to roughly 18% of the world’s population but only about 2% of global air traffic. People assume closing that gap means more aircraft. It also means keeping today’s aircraft flying: older fleets, long parts lead times, manufacturer support far away. Spotting a problem two weeks early is the difference between a planned repair and a week of cancellations.
Suppose AirVara works at the scale you imagine. What changes — for an aircraft operator, for a passenger, and ultimately for mobility across the continent?
In an ideal AirVara scenario, operators plan maintenance instead of reacting to it. Passengers get flights that leave on time. And regional aviation becomes reliable enough for businesses, supply chains, and communities to build on.
What is the most important thing you need to prove over the next 12–18 months?
The most important goal over a year is for AirVara to produce more indirect outcomes, not just proactive dashboards. One operator, one maintenance problem, one measurable result. Then repeat it and show increased utility (like handling procurement) for operators.
And what could the right partner unlock for AirVara right now — whether that partner is an aviation operator, technology company, investor, technical expert or funder?
The technology isn’t the bottleneck. Access is. We’re already working with one design partner, and every operator that shares its engine data makes AirVara’s early warnings sharper. So the most valuable partners right now are more regional and turboprop operators ready to run a focused evaluation on one agreed problem. Beyond that, MROs and data providers could cut onboarding from weeks to days, and funders could expand regional scope. If you lead maintenance at a regional operator, let’s talk!
Finish this sentence: “I’m building AirVara because…”
I’m building AirVara because the flight operators legacy systems skip are the ones keeping the world connected.
Signals before setbacks.
Conceptual illustration of a change in engine readings across flights; not actual AirVara or fleet performance data.
Leroy Mwasaru
Leroy Mwasaru is the founder of AirVara and a Prometheus X Venture Fellow. He began building technology as a teenager in Kenya, co-creating a system that converted human and organic waste into biogas. His subsequent work has spanned software engineering, education technology, intelligent automation, and aviation. Today, he is applying that systems-building mindset to one of the less visible but essential layers of African infrastructure: keeping aircraft operational.