Two 23-year-olds from rural France beat Google DeepMind on an industry benchmark within 40 days, backed by six unicorn founders, including Hugging Face and SumU.p
Minitap, the AI-powered mobile development platform, announced it has raised $4.1 million in seed funding co-led by Moxxie Ventures and Mercuri, with participation from EWOR, Tekton Ventures, Amigos Venture Capital, and six unicorn founders. The round comes just four months after founders Nico and Luc, both 23, achieved the #1 position globally on AndroidWorld, the industry benchmark for AI-controlled mobile devices, surpassing research teams from Google DeepMind, ByteDance, Microsoft Research, and Alibaba.
Mobile development is 10x slower than web development, even in the age of AI. While tools like Cursor and Claude enable web developers to ship features in 2 days that previously took 2 weeks, those same AI tools remain largely ineffective for mobile—unable to test on devices, iterate when things break, or verify features work across configurations. Minitap solves this bottleneck, enabling engineering teams to build mobile features in days, instead of the usual six weeks.
AI Authority Trend: Ostorlab Launches AI Pentesting Engine to Transform Mobile App Security
The Real Story Behind the Milestone
The headlines center on two 23-year-olds from rural France who defeated Google DeepMind in 40 days; however, the real story is about the shift of AI innovation from academic labs to product teams that are driven by execution and solving problems that they have personally experienced. Minitap did not win by creating a bigger model; it won by creating the missing layer that everyone else overlooked: infrastructure that allows AI to be able to test, break, learn, and fix on real mobile devices at a large scale.
AI tools have been transforming web development for several years, while mobile has remained very slow, not because of lack of intelligence, but because AI did not have the sensory access and feedback loop that humans rely on when debugging apps. By giving AI the capability to interact with phones like a real engineer – i.e, “hands and eyes” – Minitap has changed mobile development from a sequentially human-dependent process into one that is powered by autonomous iteration.
The AndroidWorld triumph is not only a benchmark win; it is a transition to a new paradigm where the teams that are leading the future of software are not necessarily the largest ones, but those that are willing to build vertically, face reality directly, and give AI agency instead of assistance.
“We spent two years building our first viral mobile product, today, and I’m embarrassed by that timeline,” said Nicolas Dehandschoewercker, co-founder and CEO of Minitap. “Mobile is 60% of internet usage, but moves at 10% of web speed. Every consumer app company (Duolingo, Calm, Hinge etc) ships 5x more experiments on web than mobile. We built Minitap to close that gap for everyone.”
“Every consumer mobile company needs to experiment faster,” added Luc Mahoux-Nakamura. “The companies that run 10x more experiments will win their markets. We’re building the infrastructure that makes that speed possible.”
Nico and Luc met in Cosne-Cours-sur-Loire, a village in Burgundy, France. Nico had spent two years in military school; Luc was a child prodigy. Luc taught Nico how to code, and they studied together every day to rank in France’s top 0.1% academically. Seven years later, they’ve built every project side by side.
At 18, they created their first mobile app and bootstrapped Fuego to 10,000 users. Nico went on to study Biomedical Engineering at Imperial College London and carry out AI research, obsessed with DeepMind’s AlphaGo documentary. Luc built delivery drone infrastructure at Rakuten. Together, they combined mobile development experience, AI research expertise, and infrastructure built at scale, a combination no research lab could match.
“Nicolas is leading one of the fastest teams I’ve seen. It comes from years of working together, knowing mobile inside out, and understanding how to build AI systems that hold up. The combination of AI research capabilities, mobile development skills, and sheer hunger of will is unprecedented and ideal for solving this specific problem,” said Daniel Dippold, Founder and CEO of EWOR.
Minitap’s technical achievement centers on two innovations: mobile-use, an open-source framework that lets AI agents control phones like humans, and minitap cloud, infrastructure that instantly spins up any phone configuration – iOS or Android – across thousands of devices in parallel. These tools connect to AI coding environments, enabling AI to write mobile code, test it on real devices, identify bugs, fix itself, and ship working features autonomously.
Within their first 40 days, Minitap claimed the #1 position on Google DeepMind’s AndroidWorld benchmark, the industry standard for measuring AI control of mobile devices. The founders then open-sourced their entire solution, advancing the field and growing their repository to 1,900 GitHub stars.
AI Authority Trend: Cloudflare Acquires Replicate to Build Global AI Development Platform
The funding round attracted a high concentration of unicorn founders and AI infrastructure experts: Thomas Wolf (Hugging Face, $4.5B valuation), Stefan Glanzer & Michael Breidenbrucker (Last.fm), Paul Muller (Adjust, >$1B exit), Petter Made (SumUp, $8B valuation), Daniel Krauss, Jochen Engert & André Schwämmlein (FlixBus, $3B valuati, on)and Saturnin Pugnet (Worldcoin). The round also includes operators from OpenAI, DeepMind, LangChain, and LlamaIndex.
“When you see two 23-year-olds from rural France beat Google in 40 days, you recognize something rare,” said Katie Jacobs Stanton, Founder and General Partner at Moxxie Ventures. “Nico and Luc are solving a massive problem that they uniquely understand and are moving at an urgent speed.”
Today, engineering teams at consumer mobile companies use Minitap to build features 10x faster. Minitap aims to enable growth teams to ship features without going through the engineering. A product manager describes a feature, provides a Figma design, and AI generates code, tests it, and ships an A/B test in one afternoon.
Longer term, the team plans to build mobile apps that optimize themselves autonomously, running experiments, analyzing user behavior, generating hypotheses, building variations, measuring results, and iterating—all without human intervention.
“Minitap is one of the first companies that is bringing agentic AI to mobile use and possibly the very first that is taking a full stack approach to enable the use of AI coding agents for mobile app development. This is a substantial challenge and a huge opportunity that Nico and Luc are uniquely positioned to solve” said Esha Vatsa, Partner at Mercuri.
Minitap is an AI-powered mobile development platform that enables engineering teams to ship features 10x faster. Their vision is to make mobile development so seamless, non-technical teams at consumer app companies can build entire features on mobile apps without going through the engineering. Founded in 2025 by Nico and Luc, Minitap combines open-source AI agent frameworks with cloud infrastructure to automate mobile testing and iteration. The company achieved #1 globally on AndroidWorld, the industry benchmark for AI-controlled mobile devices, and has raised $4.1 million from Moxxie Ventures, Mercuri, and six unicorn founders.
AI Authority Trend: Onton Raises $7.5 Million to Reinvent How People Discover and Buy
FAQs
1. How can Minitap feasibly cut down the mobile development time from six weeks to a few days?
Simply, it eliminates the part of the development cycle which is the slowest by far – the waiting time when humans are debugging what has gone wrong in reality. Minitap enables AI to create code, put it on real devices, see the results, find the problems, correct itself, and repeat the process. While human engineers work sequentially, AI agents can work in parallel across thousands of devices. So, the speed does not come from the AI “coding faster,” but from the fact that the human-paced feedback loop which used to define mobile development, is no longer there.
2. Non-technical teams are capable of feature deployment without the intervention of engineers, right?
Yes, but it is not done in a gimmicky “no-code” manner. It is more a shift in the structural workflow. Instead of fighting for the engineering resources to conduct experiments, a product manager gives a feature description and a Figma design, and Minitap creates the code, tests it, and completes the A/B experiment without any help. The growth becomes dependent on creativity and insight rather than developer hours. The teams no longer “wait for resources” but rather think, “How many experiments can we come up with?”
3. What prevented AI tools like Cursor or Claude from having similar impacts on mobile development as they did on the web?
The difference is that web development takes place in a predictable environment where AI can reason abstractly, whereas mobile is affected by the real-world hazards of device fragmentation, build failures, OS inconsistencies, and unpredictable edge-case bugs. Previous AI tools could produce code, but they failed when non-linear debugging for mobile was presented. Minitap’s major innovation is that it provides AI with the capability to interact with reality, fail, adjust, and repeat the process – thus, the autonomy loop which was missing is now closed.
4. What does outperforming Google DeepMind on the AndroidWorld benchmark really mean?
It is much more than just winning a leaderboard. The outcome reflects that the cutting-edge AI work is moving from simple large-scale model to more agent-based system – systems, with the ability to test, correct, and persist. A small team in 40 days accomplished what big labs haven’t: the groundwork for AI to work like a full-stack mobile engineer. The benchmark was not only about performance but also control, resilience, and problem-solving through iteration, and Minitap was leading, where just being big was not enough.
AI Authority Trend: ISG Says: GenAI Transforms Application Development in Germany
To share your insights, please write to us at info@intentamplify.com
