loading...
DeepHow lets skilled workers record themselves narrating procedures on a phone, then uses AI to slice footage into step-by-step lessons with overlays and automatic translations.
ENTRY ANGLES
Point-of-work guidance systems (just-in-time instructions vs. traditional training curriculum) · Enterprise-focused platform targeting large manufacturers with thousands of workers · AI-integrated training platforms for worker instruction
VERTICALS
CAPABILITIES
AI capabilities for content generation and worker guidance, Enterprise sales and customer success for large manufacturers
DeepHow was [covered previously](/review/est-kogo-uchit-krome-ajtishnikov) in late 2021. Since then it has raised two additional $14M rounds in quick succession, bringing total investment to $37.1M – a pace that signals something is working.
The platform trains industrial workers through video. Its core workflow: experienced workers record themselves performing procedures on mobile devices, narrating their actions as they go. DeepHow uses AI to slice that footage into step-by-step lessons, embed text overlays and diagrams, and generate translations automatically. The finished lessons go into a searchable catalog accessible on phones or tablets mounted near equipment.
The newest version adds GPT-class capabilities. Upload a PDF procedure document and DeepHow generates a filming script – scene breakdowns, narration cues, everything. Shoot the video and the platform auto-creates a written instruction and process flowchart. A natural-language chat interface lets workers ask questions instead of keyword-searching; the system returns the relevant lesson instantly.
The efficiency gains are concrete: 25% reduction in training time, 40% faster ramp to full productivity for new hires, 20% less time from subject-matter experts on shop-floor problem-solving. DeepHow has over 70 enterprise clients, including ABInBev, Siemens, Stanley Black & Decker, Energizer, Nissan, and Komatsu. Revenue has grown 3x year-over-year for several consecutive years.
The skilled-labor shortage conversation almost always defaults to software developers and data scientists. The industrial workforce tells a different story. By 2020, finding a qualified industrial worker was already 50% harder than it had been two years earlier, and the gap has kept widening. Roughly 2 million industrial positions over the 2020–2030 decade are expected to go persistently unfilled due to qualification mismatches.
The scale context matters: there are 700 million industrial workers globally. A [related review](/review/programmirovat-nuzhno-ne-kompjutery-a-ljudej) put that number alongside the roughly 28 million developers worldwide. Any edtech platform targeting the developer market is pursuing a fraction of the addressable audience DeepHow is working with.
High turnover compounds the training burden – companies aren't just onboarding new workers, they're re-onboarding constantly. The workforce training platform market for skilled trades was $17.1B in 2021 and is projected to reach $62.9B by 2030, growing at 16.2% annually. Comparable platforms active in the space include how.fm (warehouse workers, $8.1M raised), BuildWitt (construction workers, $7.6M), Kognitiv Spark (AR-assisted training and remote problem-solving, $3.9M), and Zaptic (digital work instructions, $16M).
Building training platforms for industrial workers is the obvious direction, but the frame matters. The platforms that gain traction are not the ones that digitize classroom theory – they're the ones that deliver just-in-time instructions exactly where a worker needs them, at the moment of need. That shift from "training curriculum" to "point-of-work guidance" is the design insight worth replicating.
DeepHow's enterprise focus and aggressive AI integration are the two specific moves worth copying. Selling to large manufacturers with thousands of workers produces the revenue growth rates that attract follow-on capital. AI capabilities aren't just a feature – they're the main argument for replacing whatever the factory currently uses. In the current investment climate, the AI angle also significantly improves fundraising leverage.