He arrived in America on Christmas Day with a barcode-shaped view of the world. Then he found the trillion-pound problem that barcodes never solved - and built Rebulk to fix it.
The barcode is one of those inventions so complete that most of us forget it was ever invented. Scan a can of soup, a pallet of laptops, a shipping container full of sneakers, and a computer somewhere knows exactly what moved and where it went. But walk out to a farm, a rail yard, a mine, or a biomass storage site, and the barcode's tidy logic falls apart. Feed sits in a pile. Grain fills a bay. Sand spreads across a yard. Logs stack in a heap. None of it fits in a box, and so almost none of it gets counted with any real accuracy. That gap - the difference between what's on a spreadsheet and what's actually sitting on the ground - is where Warren Wang decided to build a company.
Wang is the co-founder and CEO of Rebulk, a startup in Olathe, Kansas that graduated from Y Combinator's Summer 2025 batch. The pitch is deceptively plain: point cameras and LiDAR at a pile of stuff, run computer vision over it, and produce an accurate, live measurement of how much is there. Rebulk calls itself the source of truth for bulk operations. In practice, that means turning the guesswork of industrial inventory into numbers a team can actually trust.
It is not a glamorous problem. It is a huge one.
Most bulk inventory today is managed with rough estimates, manual checks, and delayed updates. Someone eyeballs a pile. Someone climbs a silo. Someone runs an expensive survey once a quarter and hopes the number holds. The errors compound quietly until they don't - and then they show up as shrink, stockouts, botched planning, and the kind of last-minute scramble that operators call a fire drill.
Wang did not read about this problem in a deck. He ran into it. As a product leader at Milk Moovement, a Series A agtech company that helps move roughly 45% of California's dairy and a fifth of the nation's, he worked closely with bulk feed suppliers and the managers trying to plan logistics around inventory nobody could pin down. The complaint kept surfacing in different accents from different corners of the industry: we don't really know how much is out there.
Before agtech, he had seen the operational underbelly of a very different giant. At Chick-fil-A's corporate support center, Wang worked in strategy and operations and helped launch autonomous ground robot and drone delivery programs - a front-row seat to what happens when you try to bolt new technology onto messy, physical, real-world logistics. Then he spent time on the other side of the table entirely, as an investor at Dynamo Ventures, backing early-stage industrial and supply chain startups. He had studied the space as an operator, as a technologist, and as a check-writer before he ever wrote his own first line of company.
Rebulk did not start as Rebulk. It started as dScribe AI, a video transcription tool, born in 2024 after Wang met his co-founder Cole Robertson at a Startup Crawl event in downtown Kansas City. Robertson brought a hard engineering resume - Engineer #3 at Tropic, where he helped scale from seed to Series B, plus aviation work at Garmin and a stint running engineering at an AI startup. Two builders, one chance meeting, and a company was underway.
The transcription product worked. It just wasn't the point. As Wang and Robertson talked to more people in industrial settings, the same bulk inventory complaint kept coming back - louder and more urgent than anything transcription was solving. So they did the thing that sounds obvious in hindsight and terrifying in the moment: they killed a working product to chase a bigger problem. dScribe AI became Rebulk, a name suggested by Kenneth Bautista, who joined as chief of staff.
The pivot is the part of the Rebulk story that other founders will want to steal. It is easy to abandon something that is failing. It is much harder to walk away from something that is quietly succeeding because you can see a larger opportunity behind it. Wang's edge was that he had heard the real problem so many times, from so many operators, that ignoring it started to feel like the actual risk.
The clearest sign that Rebulk is onto something is who is already testing it. The company has launched pilots with Cargill and Charm Industrial, measuring biomass inventory across real-world storage environments - not lab conditions, but the messy, dusty, weather-beaten places where bulk materials actually live. Landing enterprise pilots at that scale as a company of a handful of people is not normal. It is the kind of early validation that gets investors to lean forward.
And they have leaned forward. Rebulk's backers include KCRise Fund, Abstraction Capital, Flyover Capital, Redbud VC, and EquipmentShare - a roster that blends Midwestern industrial credibility with startup capital. The company was named a Kansas City Startup to Watch in 2026 and selected for LaunchKC's 2026 cohort. It is, notably, building this from Olathe, Kansas, with a second office in Kansas City, Missouri - close to the feedlots, rail yards, and storage sites it serves rather than to the nearest venture hub.
The ambition stretches past measurement. Rebulk's platform now spans a bulk inventory system, a transload operating system that runs the process from railcar arrival to invoicing, track-and-trace for railcars with live ETAs, and AI-assisted switch planning for rail yards. The through-line is the same: take the parts of a bulk operation that live in someone's head or on a whiteboard and turn them into clean, connected, billing-ready records.
There is a detail in Wang's story that is hard to invent. He emigrated from Indonesia to the United States on Christmas Day 2016. And on the same day his company announced its rebrand to Rebulk, he completed his U.S. naturalization - officially taking his full legal name, Warren Wijaya Wang. A new company name and a new citizenship, stamped on the same date. It is the kind of coincidence that reads like a screenplay note, except it happened.
You can trace a pattern through Wang's resume that explains Rebulk better than any pitch. Chick-fil-A taught him that real-world logistics resist clean software. Milk Moovement showed him exactly where the bulk inventory pain lived. Dynamo Ventures trained him to recognize which industrial problems are big enough to build a company around. Rebulk is what happens when someone who has seen a problem from three angles finally decides to solve it himself.
It also explains why he was willing to build in Kansas rather than California. The customers are here. The piles are here. When you are trying to convince a feedlot or a biomass site to trust a camera over a clipboard, being a short drive away matters more than being a short walk from a venture office.
What Wang is really betting on is that the boring, physical, uncounted parts of the economy are ripe for the same quiet revolution the barcode brought to everything else. Grain, feed, sand, logs, biomass - trillions of pounds of material that the digital world has mostly ignored because it doesn't sit still and doesn't come with a label. If Rebulk can put a reliable number on all of it, the fire drills get shorter, the planning gets sharper, and the pile stops being a mystery. That is the whole idea. It just took someone who had studied the problem from every side to see it clearly.