The AI That Listens to Factories: Scope AI and the Inspection Industrys Reckoning

When Troy Stewart, Global Director at TUV Rheinland, tried Scope AI's inspection platform, his team's reporting process went from days to minutes. He is not the only one. According to Scope, inspectors from seven of the world's ten largest inspection companies now use its software. That claim, backed by a named reference on Scope's own website from a top-three TIC provider, is the closest thing to independent verification the company has offered publicly. It is also why Index Ventures just led a $20 million Series A round into a London startup founded in 2024. Scope was built by Jonathan Low and Jakob Cassiman, two outsiders to the testing, inspection and certification industry. Low previously worked at Conjecture on document processing; Cassiman was a machine learning engineer focused on physical data. They met at Entrepreneur First in London, spent months cold-calling TIC firms and shadowing inspectors on site, and built the product they wished existed. "We talked to an inspector who carried a heavy manual of standards she had to reference during every inspection," Low told Generative Group. "After trying the tool for the first time, she said, 'Wow, I don't need this anymore' and threw the book away." The pitch is simple. Industrial inspections are vital to keeping planes, refineries, and infrastructure running safely. A four-hour on-site inspection can require ten days of post-field data processing, report writing, and form-filling using software that has not been meaningfully updated in two decades. Inspectors spend more time at desks than in the field. Scope's AI records findings by voice, video, or structured input, pulls relevant historical context, and generates a finished report on site. The company says this cuts reporting time by a factor of ten and reduces error rates by 95 percent. On its own website, Scope claims a 55 percent reduction in end-to-end inspection time. The variation between "ten times faster reporting" and "fifty-five percent faster overall" is not reconciled in any public document. What makes this round notable is not the size. Twenty million dollars is modest by AI standards. The notable thing is the market. The global testing, inspection and certification industry generates somewhere between $250 billion and $300 billion in revenue annually, according to different market research estimates. The major players, SGS, Bureau Veritas, Intertek, TUV SUD, and DEKRA, collectively hold most of the market and have for decades. Their software stacks are legacy. Their inspectors are aging. Scope says 40 percent of key technicians in the industry are expected to retire within the next five to ten years; roles take ten years to qualify for and an average of ten months to fill when someone leaves. The company was not formed to displace these workers. It was formed to do more with what remains. Stephane Kurgan, the Index Ventures partner who led the round, put it this way in a statement: "Jonny knows the intricacies of the TIC industry inside out. He is spending an incredible amount of time in the field to understand inspection problems in a granular way, and feeds them back to the product team." The investor framing treats this as founder obsession solving a specific vertical pain. The underlying bet is that TIC firms facing a workforce cliff will pay for software that makes each remaining inspector dramatically more productive, rather than trying to hire their way out of the problem. The structural logic is coherent. Whether the numbers hold is another question. All of Scope's published performance claims are company-sourced. The 100 percent pilot conversion rate, the 9x ARR growth since July 2024, the error reduction figures: none have been audited or independently verified. The named customer quotes on Scope's website are endorsements of the product, not independent benchmarks. The market size and retirement statistics cited by the company also lack third-party corroboration. What exists publicly is a company making claims about itself, backed by an investor who led the round. The company's own website, published May 21, lists three named customer testimonials: Chris Hassell, Head of Inspection at NDT Group; Troy Stewart, Global Director at TUV Rheinland; and a Global COO at a top-three TIC provider who requested anonymity. All three describe meaningful productivity gains. None constitute independent analysis. The underlying question, whether Scope's AI has genuinely compressed a ten-day post-inspection process into twelve minutes at scale across different inspection types, has not been tested by anyone outside the company. What Scope is building toward is larger than report writing. Its stated vision is to turn every inspection into a data point in a live, queryable database of physical infrastructure, training models on how assets behave, degrade, and fail. If that vision matures, the company would own the inspection knowledge layer for a significant portion of industrial infrastructure. The incumbents built their businesses on inspector expertise and accreditation frameworks. Scope is betting it can build the data layer underneath both. The pressure on incumbents is real and growing. The question is whether this particular startup is the one that captures it, or whether SGS, Bureau Veritas, or one of the other giants builds or buys the equivalent capability first. Scope has roughly twenty months of commercial traction, a $20 million war chest, and a first-mover claim in a market that has not seen serious software competition in two decades. That is enough to watch.