In the world of physical machinery from elevators and vending machines to complex mining equipment a critical resource is quietly disappearing: tacit knowledge. When a technician retires, years of on-the-job expertise, those undocumented tricks and solutions often vanish with their legacy. Companies are left scrambling with a skill gap and expensive, prolonged training for the next generation.
Enter Integuide.
Integuide, founded by Hunter Jay and Slava Chalnev, is building AI to guide technicians through maintenance and servicing tasks. The goal is simple: to make the deep, real-world knowledge trapped inside a technician’s head immediately accessible to every new tech in the field. They are figuring out how to apply the power of modern AI to the messy, non-digital world of physical infrastructure.
From Fruit Bots to Physical AI: The Genesis
Hunter and Slava aren't newcomers to the deep end of AI. Hunter, an AI engineer, previously built a company focused on fruit-picking machines and has been engaged with AI for eight years. Slava, also an AI engineer, spent years as an interpretability researcher, trying to figure out how language models work on the inside.
AI has already proven to be profoundly transformative in many industries. Integuide is seizing this opportunity applying AI beyond digital screens and into physical, hands-on environments that need a leg up.
They met with the shared belief that they needed to do something meaningful in this space. Hunter sums up their broader mission: "We're trying to apply AI into real-world tasks."
The Pivot That Unlocked the Problem
Integuide started with a logical, yet incomplete, idea: use AI to analyse a machine's thousand-page documentation. But as they talked to customers, they hit a wall.
"It turns out that manuals don’t contain all the information you need," Hunter recalls. "You have to record the actual work being done, which is obvious in hindsight."
The truth was, the most valuable knowledge was not documented explicitly as it was in the minds of the experienced technicians.
How Integuide Works Today
- Low-Effort Capture: Integuide provides a small body camera to the technician. When they perform a 500-hour maintenance or troubleshoot a rare fault, they simply record themselves. This is the low-effort way to capture tacit knowledge that would otherwise take days to document manually.
- The Analysis Engine: The recorded video, along with any existing manuals, is fed into Integuide's AI analysis engine. This engine breaks down the giant set of messy data, understands the steps taken, and synthesises the core procedure.
- The Output: The AI transforms the raw data into usable assets such as:
- Ideal Operating Procedures: A clean, step-by-step document that synthesises the best practices, even if ten different recordings showed different approaches.
- Interactive Q&A: The knowledge base becomes a Q&A assistant. No more senior techs getting 12 calls a day from junior staff; the first stop is the AI.
This is a massive uplift to process documentation, particularly in the hardware world. As Slava notes, "The AI does the heavy lifting of the cleaning it up and making it nice."
What’s next for Integuide?
Slava posits a future where they have "figured out what this new form factor of organisation centered around AI in physical spaces looks like." The ultimate horizon? Full robot automation, guiding machines to do the tasks themselves.
They are convinced that as AI gets a better understanding of the physical world, there's a "blob of intelligence that we're trying to figure out how to best enable and best structure" to get useful work out of it.
For companies dealing with complex machinery and the constant headache of training, Integuide is building a knowledge bridge that ensures the hard-earned expertise of today doesn't become the skills gap of tomorrow.
Get in touch with the team!
Integuide is currently looking for introductions to companies doing physical work, especially those with technicians working on equipment in:
- Lifts & Vending Machines
- Construction Equipment
- Electrical & Mining
- Transport
If you or someone you know is trying to solve the problem of knowledge transfer in a physical setting, the future of maintenance might be a body cam away.
