Tunisian Founders Who Sold For USD 120 M Raise USD 9 M To Do It Again

Deal Street

African Startup Funding Tracker

US$741,804,000+

*Data updated daily at 18:00 EAT

Tunisian Founders Who Sold For USD 120 M Raise USD 9 M To Do It Again—With AI

Eighteen months ago, Karim Jouini and Jihed Othmani were ready to retire from startup life, fresh off a nine-figure exit.

Their expense management platform, Expensya, had just been acquired by Swedish fintech Medius in a deal reportedly worth over USD 120 M—one of Africa’s largest tech acquisitions made in Tunisia.

But the pull of generative AI and a nagging sense that they had unfinished business has drawn them back into the ring.

Their new startup, Thunder Code, has raised USD 9 M in seed funding to automate and rethink software testing from the ground up using generative AI.

Led by Silicon Badia, with participation from Janngo Capital, Titan Seed Fund, and strategic angels like Roxanne Varza of Station F and Karim Beguir of InstaDeep, the round includes familiar names from the Expensya era, some of whom are former employees turned investors.

Thunder Code is betting that quality assurance (QA), an often-overlooked but crucial bottleneck in software delivery, is ripe for reinvention.

The startup’s platform uses AI “agents” to autonomously understand apps, generate and execute tests, and catch bugs, promising to cut testing time by up to 90%.

In a world obsessed with shipping faster, it’s a pitch that’s already gaining traction with pilot programs in the U.S., France, Tunisia, and Canada.

Unlike Expensya, which took years to mature, Thunder Code shipped its MVP in just six weeks. “We’re moving 10x faster this time,” Jouini says, noting that the product today is already more robust than Expensya was in year four.

The founder emphasises that from day one, they have applied hard lessons: ship fast, hire top-tier talent early, and don’t be afraid of dilution if it buys speed and expertise.

Their timing is sharp. The global software testing market is projected to top USD 100 B by 2027, yet much of it still relies on clunky, code-heavy platforms.

Thunder Code joins a growing list of startups racing to modernise testing with AI, from incumbents like Tricentis to new entrants like Nova AI, but believes its execution speed and real-world traction give it a meaningful edge.

More than just a second act, Thunder Code feels like a startup born from unfinished ambition. “We promised not to do this again,” Jouini admits. “But the opportunity felt too big to ignore.”

← Previous Article