Walk into almost any small or mid-sized factory in America and you’ll find the same thing: decades of data, and almost no intelligence. The machines generate readings. The orders pile up in spreadsheets. The most important knowledge of all, how things actually get made, lives in the heads of the people who’ve worked the floor for twenty years. It’s all there. Almost none of it is usable. That gap is the whole reason Corello exists, and it’s the sentence Carlos Maiguel kept coming back to in our conversation: the data exists, and the intelligence does not. This episode of The Corridor is about what happens when you point modern AI at one of the least glamorous, most essential industries there is, and why the factories the tech world overlooks may be exactly where AI creates the most value. Why manufacturing, of all places Carlos didn’t stumble into manufacturing. He chose it, deliberately, after spending years in the trenches understanding it. While most of the AI world races toward consumer apps and developer tools, he went the other direction, toward small and mid-sized manufacturers, the businesses that actually make the physical things the economy runs on, and that have been almost entirely skipped by the software revolution. That’s not a coincidence. It’s the opportunity. These factories run on legacy systems, paper, tribal knowledge, and instinct. The bigger players have the budgets for expensive enterprise software; the small and mid-sized ones have been left to fend for themselves. Which means the ceiling for improvement, in efficiency, in speed, in quality, is enormous, precisely because no one has bothered to build for them. Data everywhere, unused The core problem Corello solves is one Carlos describes simply: data is everywhere, but it’s fragmented and unused. A factory’s information is scattered across machines, systems, spreadsheets, and people, none of it talking to each other. On its own, that data just sits there. It doesn’t answer questions. It doesn’t catch problems. It doesn’t help the person on the floor make a better decision in the moment. The raw material for intelligence is present in abundance; the intelligence itself is missing. Corello’s work is to ingest that fragmented data and turn it into something usable, to close the gap between what a factory already knows and what it can actually act on. That’s the difference between a factory drowning in data and a factory that can finally use it. The knowledge that walks out the door The most human part of the problem is tribal knowledge, and it’s the piece that should worry every factory owner. In these businesses, the most valuable knowledge isn’t written down anywhere. It lives in the experienced worker who knows, by feel, how to set up a machine, spot a defect, or quote a complex job. When that person retires or leaves, decades of hard-won expertise walk out the door with them. Nothing captures it. Nothing preserves it. The next generation starts closer to zero than they should. This is where AI becomes genuinely powerful, not as a replacement for those workers, but as a way to capture and preserve what they know. Carlos frames the goal not as automation that removes people, but as AI coworkers, goal-driven team members that work alongside the humans on the floor, carrying the institutional knowledge and helping everyone operate at the level of the most experienced person in the building. The expertise stops being fragile. It becomes an asset the whole factory can draw on. Building AI-native systems in old industries The hard part, and the reason few people attempt this, is that building AI-native systems on top of old industries is genuinely difficult. You’re not deploying software into a clean, modern tech stack. You’re integrating with legacy systems, decades-old machines, and workflows that were never designed for digital tools. And you’re doing it for customers who, understandably, need to see real return before they trust something new on their floor. That’s why Carlos emphasizes ROI and speed. A factory owner doesn’t care about the elegance of the technology; they care whether it saves time, reduces waste, and makes them money, fast. And trust has to be earned on the floor, with the people doing the work, not just in the front office. Corello’s use cases, things like faster quoting and clearer work instructions, are deliberately practical: real problems, real time saved, real value the moment it’s turned on. The lesson underneath What makes this episode matter beyond manufacturing is the founder philosophy underneath it. Carlos’s approach to building, understand the problem deeply before you build anything, and build openly by leaning on your network, is the same discipline that makes any hard company work. He spent years understanding his customer before writing a line of product, and he’s built in the open, gathering feedback and drawing on the community of founders who came before him. There’s a bigger point here too, one that sits at the heart of why this show exists. The most valuable opportunities are often in the unseen spaces, the industries too unglamorous for the spotlight, the customers the tech world forgot, the founders the market overlooks. Carlos is building for all three at once. And the reason that’s a smart bet, not a charitable one, is simple: the data was always there. Someone just had to build the intelligence. Listen to the full episode of The Corridor with Carlos Maiguel. Seguimos. 🇵🇷 Angel León, Coquí Ventures Coquí Notes is the editorial publication of Coquí Ventures, the diaspora platform connecting Latino founders to community, capital, and the infrastructure that turns ideas into category-defining tech companies. Get full access to Coquí Notes at coquinotes.substack.com/subscribe