Shaping the Future of Software Testing
Testing shouldn't be the bottleneck. It should be the reason you ship with confidence. That's what we build at Thunders.
fueled by experience
Our Story
We've lived the pain. At Expensya, we encountered test files with over 10,000 lines of code; we know exactly what it feels like when testing becomes unmanageable. Flaky tests, endless maintenance, processes that grind progress to a halt. After twenty years of shipping software, from Microsoft to Expensya, we sat down in late 2024 and asked ourselves a simple question: "What if AI could handle the grunt work, freeing teams to focus on what actually matters?" That question became Thunders. A company on a mission to make testing faster, smarter, and built for humans.
2008
Karim Jouini joins Microsoft as a software engineer, shipping products to hundreds of millions of users
2012
Jihed Othmani, Karim's longtime friend, joins Microsoft. Both spend their evenings taping receipts to paper and retyping them into spreadsheets, and wonder why nobody has fixed this
2014
Karim and Jihed found Expensya in Tunis and Paris to solve that pain. Over the next decade it grows to 160 people across 20 countries, 6,000 customers and 700,000 users in 100+ countries
2021
Expensya raises a $20M Series B
2023
Expensya acquired by Medius for over $100M. Karim becomes Medius Chief Product & Technology Officer, Jihed its Platform CTO, integrating six companies across three continents
Q2 2024
Twenty years, millions of lines of test code, hundreds of engineers managed. Across every project, Karim and Jihed kept running into the same wall: testing. Slow, brittle, always the last thing standing between the team and the release, and no tool on the market ever fully solved it. So they did what they did in 2012. They started building the solution themselves
Q4 2024
- Company concept developed by Karim Jouini & Jihed Othmani - Thunder Code SAS incorporated in Paris, November 2024
Q1 2025
- Thunders launches between Paris and Tunis, January 2025 - MVP shipped in week six - Listed on Microsoft Azure Marketplace - Selected for Microsoft GenAI Studio
Q2 2025
- First paying clients across the US, Canada, France and Tunisia - VivaTech presence - $9M seed round led by Silicon Badia, with Janngo Capital and Titan Seed Fund
Q3 2025
- Team grows to 15+ across Paris and Tunis - Thunder Code becomes Thunders - New website; platform opens to self-serve signup - ISO 27001 and SOC 2 Type 2 certified, July 2025
Q4 2025
- Named to STATION F's Future 40 2025 - Selected for the Meta and Allianz accelerator programs - Named "Entreprise IA de l'année 2025"
Q1 2026
- One test across several apps: admin panel, web app and API in a single flow - Data-driven testing, Jira and Linear issues turned into tests - Thunders MCP: create and run tests from Claude, ChatGPT or Cursor
Q2 2026
- Thunders North America, Inc. opens in Boston, April 2026: our first US base, closer to the teams we already serve in North America - Analytics, run comparison and shareable runs ship - Mobile testing launches
Q3 2026
- Thunders Copilot becomes Thunders AI - iOS and Android testing on real devices and emulators - Self-hosted releases move to calendar versioning, shipped in step with Cloud
15+ ISTQB Certifications across the Team
Deep expertise in both testing methodologies and cutting-edge AI
Three offices: Paris, Tunis, Boston. A remote-first team.
who know the field
Our Team
Our team is a global blend of QA veterans, AI/ML innovators, and product builders who've been in the trenches. We understand the struggles of testing because we've lived them. Every feature, every solution, every line of code is born from real-world experience and a drive to help teams everywhere.
How we Think
AI that delivers, not just promises
We're not chasing hype. We build one platform for end-to-end testing: AI test agents that reduce manual work, catch hidden bugs, and speed up releases. Less time firefighting, more time creating.
Quality without compromise
Reliability, security, and thorough coverage are the baseline. Testing is never the weak link in your product chain.
Testing for everyone
Testing should be accessible, not intimidating. Plain-language tests let QA engineers, developers, and product teams work from the same suite, with the evidence to back every result.
Built with, not just for
Our best ideas come from the people who use Thunders every day. We listen, iterate, and ship. Every improvement starts with a real-world problem.
Why Thunders, when any AI can write a test?
It can. Ask Claude or ChatGPT for a Playwright script and you'll have one in five minutes. Then the real work starts: keeping it alive through every release, running it somewhere, knowing why it failed, proving to an auditor that it ran at all. Five minutes saved on writing, hours spent on everything after.
That's the gap we close. Thunders is not a faster way to write scripts. It's the platform that gives AI-generated tests what they're missing: stable execution, self-healing when the UI moves, evidence on every step, and an audit trail you didn't have to build. Generation is the first step. Everything after it is where quality lives.
Code or no code is the wrong question
Plain language on the surface for the PM who owns the release. Playwright import, JavaScript steps, MCP and CI hooks underneath for the engineer who gates the PR. Bring your existing suite; Thunders keeps it maintainable and scalable. We are not against your framework. We are the layer that makes it hold up in production.
Agents with opinions, not just logs
A green checkmark tells you a flow passed. It doesn't tell you the page took eight seconds, the contrast failed WCAG, or a console exception fired behind a working button. Our agents test the way a specialist would: QA, accessibility, UX and security each hold your product to their own standard and tell you why something isn't good enough.
Every result comes with proof
Screenshot, console, AI reasoning and a versioned test case on every step. When a test fails, you know where and why. When compliance asks, the evidence is already there.
Ready to ship faster with smarter testing?
