London-based Meticulous, founded by brothers Gabriel Spencer-Harper and Quentin Spencer-Harper, has raised $15m in Series A funding. Ethan Kurzweil at San Francisco-based VC Chemistry led the round, which also included Menlo Ventures.
A host of angels also invested, including former Stripe operator Lachy Groom; Poolside cofounder Jason Warner; Dropbox cofounder Arash Ferdowsi; former Adobe CPO Scott Belsky; Vercel founder Guillermo Rauch; former OpenAI engineering lead Calvin French-Owen; Hex CTO Caitlin Colgrove; and former Cursor head of engineering Jason Ginsberg.
AI agents are writing more and more code quicker than software review and testing teams can keep up, meaning developers are being slowed down by long review cycles and more bugs entering their codebase.
How Meticulous works
Meticulous’s goal is to help developers trust AI-generated code so they can ship faster. The company analyses a developer’s codebase and the intended application and then creates a checklist of every edge case, meaning a rare scenario or input that exposes a failure in logic or a risk the app will crash.
For both agents and human developers, whenever a code change is made Meticulous simulates user flows before and after that code change to test every edge case that could be affected.
Before a developer merges a code change into their company’s codebase they get a report of a sequence of thumbnails, each one illustrating a different edge case that was affected by the code change.
“One may be how the change changes a given screen for a premium user, another may be how that change affects the same screen for a normal user,” Quentin says. “As you hover back and forth over the thumbnails you can see how that screen would change for the user if the code change were to be shipped.
“In a couple of minutes or less you understand the full impact of the change — you see it before your eyes — and can therefore ship almost as fast as your AI can write code.”
No change is shipped without the input of human developers, who can push changes live themselves or ask whatever coding agent they’re using for alterations.
“Since the system is deterministic it means that we can detect if, for a given change, there are relevant cases we don’t have covered, and can warn the user and agent,” Quentin says.
Currently the company tests frontend code only, but has plans to move into testing the full stack.
5x ARR growth
Meticulous has grown annual recurring revenue (ARR) by five times in the past year, counting companies like Notion, ElevenLabs, Dropbox, Wiz and LaunchDarkly among its customers.
Quentin and Gabriel both bring engineering experience into the company. Quentin spent more than 10 years at US big data giant Palantir, while Gabriel used to be a software engineer at cloud-based storage company Dropbox.
“Back then this was pre-AI, and even then if you looked at how engineers actually spent their days, only a fraction of it was spent writing code,” Quentin tells Sifted. “Engineers spent a lot more time reasoning about code and trying to verify that it works. And the larger the application the harder that gets.
“At the same time, at this scale of complexity, there was no product or solution that even came close to solving this problem,” he adds. “And so I got thinking from first principles on how you’d solve this problem at this kind of scale and complexity, millions lines of code, hundreds of developers, tens of thousands of edge cases. Meticulous was the only way I could see it working practically.
“Now with AI this imbalance is even more extreme, engineers are spending almost all of their time verifying the coding output of AI before shipping it.”
Where Meticulous will invest
A portion of the funding round will be spent on marketing. “Historically we’ve relied solely on word of mouth,” Gabriel says, “we’re changing this, we want every company to know about Meticulous.”
The company will also invest in R&D to expand into backend and performance testing, as well as headcount growth.
The team is currently 20-strong, but the founders want to grow that to around 30 or 40 in the next 12 months. “We want to keep the team small but exceptional,” Gabriel says.
Meticulous is also openly hiring for forward deployed engineers (FDE), a role that has become central to a host of European AI companies including Mistral, ElevenLabs and Lovable. FDEs are software engineers who embed directly with a customer, often on-site, to help identify problems and solve them by deploying their company’s technology in a bespoke way.
The market for talent is “very competitive”, Gabriel tells Sifted, adding that “the importance of what we’re building” is their edge among software engineers.
