Here is a problem that sounds boring and turns out to be enormous. You have designed a physical thing - a bracket, a housing, a gear, some small precise object that a robot or a satellite or a medical device needs to exist. You have the CAD file. What you do not have is a factory. So you email machine shops, you wait, you get quotes back in a week if you are lucky, you discover the good ones want you to order 500 units when you need 4, and somewhere in this process the calendar quietly loses a month. Multiply that by every hardware startup on earth and you have described one of the least efficient markets in the modern economy.
Factorem is a Singapore company built on the theory that this market is a software problem wearing a manufacturing costume. Founded in 2020 by Alexandra Zhang and Hardik Dobariya, it is an on-demand manufacturing platform: you upload a design, its software reads the geometry, checks whether the thing can actually be made, matches it to the factory best suited to make it, and returns a quote - often instantly. Then it manages the order, the quality control, and the logistics through to delivery. The pitch is that the whole thing behaves less like a supply chain and more like an online checkout.
The founders did not arrive at this idea through market research. They arrived at it by being annoyed. Both are National University of Singapore graduates - Zhang from real estate, Dobariya from electrical engineering - and both did internships in Toronto under the NUS Overseas Colleges program, where they worked at product companies and ran directly into the sourcing wall. Producing prototypes was slow and expensive, and finding a reliable manufacturer was somehow worse. This is the origin story of a great many companies: the founders had a specific pain, decided the pain was structural rather than personal, and went to build the thing that would have made their own lives easier. NUS Enterprise incubated them at The Furnace and handed over a grant. The name they chose, Factorem, is Latin for "maker," which is the kind of naming decision that tells you the founders spent at least one evening feeling romantic about what they were building.
The mechanism underneath is more interesting than the marketing. Most factories, most of the time, are not running at full tilt. There is spare capacity - machines that are idle, hours that go unsold. In manufacturing this idle time is a peculiar kind of inventory: it is worth something, it expires continuously, and nobody has historically been able to sell it efficiently. Factorem's model is to aggregate demand on one side and route it to that spare capacity on the other. The customer gets a part faster and cheaper; the factory monetizes time it was going to waste anyway; and Factorem sits in the middle taking a cut for making the match. Marketplaces that work usually work this way - not by conjuring new supply, but by unlocking supply that already exists and was simply illiquid.
The menu is broad. Factorem handles CNC machining, 3D printing, sheet metal fabrication, injection molding, surface treatment and finishing, and re-engineering of legacy or hard-to-source parts, drawing on a catalog of more than 200 materials. The network it routes to has grown from roughly 10 factories at launch to more than 110 vetted partners spread across Singapore, Malaysia, Vietnam and South Korea, and the platform reports more than 30,000 parts manufactured to date. The customers tend to be the people building genuinely hard things: robotics teams, biotech and semiconductor firms, industrial automation, telecommunications, aerospace and defense. What they have in common is that they need a specific part, they need it right, and they need it soon - and that the cost of getting any one of those three wrong is high.
The word "vetted" is doing quiet but heavy lifting in all of this. Anyone can build a form that collects CAD files. The actual business is the boring, unglamorous, deeply defensible work of qualifying factories - checking that a partner can hold a tolerance, will hit a date, and will ship a part that matches the drawing rather than a part that merely resembles it. This is why the choice to anchor the network in Singapore is strategic rather than sentimental. Singapore does not have the largest factories in the region. It has a reputation for precision and trust, and in a business whose entire value proposition is "the part will be correct," reputation is the moat. A marketplace for custom hardware lives or dies on whether the buyer believes the promise, and belief is expensive to manufacture.
Then there is the decision to abolish the minimum order quantity, which is a bigger deal than it looks. The MOQ is one of manufacturing's oldest defensive rituals - the factory wants a run of 500 because setup is expensive and small jobs are annoying. Factorem lets you order one part, or a hundred, or a full production run, without that floor. Removing a constraint your competitors treat as a law of nature is a real strategy, because it opens the market to exactly the customers who were previously locked out: the prototyping team, the pilot run, the engineer who needs four of something this week. It is also, not incidentally, how you build habit. The team that orders one part today because it was easy is the team that orders a thousand next year.
The financing has been deliberately modest. In October 2023 Factorem closed a seed round co-led by Enterprise Singapore's SEEDS Capital and Blue InCube Ventures, with participation from Iterative and other backers; total funding reported across trackers lands in the low hundreds of thousands of dollars, which is a lean number for a company operating physical supply across four countries. That leanness is at least partly on purpose. Zhang has said that "only one or two things are truly critical to the survival of the business," which is the sort of line founders say after they have made the mistake of trying to do everything and learned to stop. The company has cited 300% year-over-year revenue growth, a figure worth treating with the usual respect one gives to fast growth off a small base - impressive, real, and not yet a business at scale.
What Factorem is really selling, underneath the CNC machining and the instant quotes, is time. The old way of sourcing a custom part measures itself in weeks and phone calls. The Factorem way tries to measure itself in days and clicks. For a hardware team, time-to-part is time-to-market, and time-to-market is very close to the whole game. If the company keeps compounding the network - one more vetted factory, one more material, one more part shipped correctly - it is building the kind of unglamorous infrastructure that nobody tweets about and everybody quietly depends on. Hardware is hard. Factorem's bet is that the sourcing part of it does not have to be.