Reframe
Systems

Inside a 17,000-square-foot shed off a quiet road, a robot is framing a wall faster than a crew could measure one. Down the line, a person picks up where the machine stops - wiring, plumbing, the human parts. This is a home being born, and nobody is pouring concrete in the rain to do it.

Reframe Systems calls this room a microfactory. The word is doing a lot of work. It is not a mega-plant with a decade-long payback and a fragile dependence on one regional housing market. It is small on purpose - cheap enough to stand up in under three months, close enough to a build site that finished modules ride a short truck instead of a long one. The Andover floor turns out roughly 45 homes a year. The plan is to copy it, not to enlarge it.

The homes that come out are all-electric, solar-ready, and built to Passive House airtightness. Some are modern takes on the New England triple-decker. Some are designed to survive a wildfire. All of them are meant to do something the housing industry rarely manages - pencil out.

Housing is expensive, slow, and oddly allergic to progress

Caption: every other industry got a productivity curve. Construction got a clipboard.

Here is the uncomfortable arithmetic. Buildings account for roughly 27% of annual greenhouse gas emissions. Construction productivity has barely moved in decades while almost every other sector compounded. Skilled trades are aging out faster than they are being replaced. And a net-zero home - the kind that would actually help on climate - has historically cost a premium that most buyers simply cannot pay.

So you get a standoff. The homes we should build are the homes we can least afford. Traditional offsite construction promised a fix, then mostly delivered enormous factories that needed enormous, steady demand to survive. When a regional market wobbled, the factory - and sometimes the company - went with it.

The founders had spent years watching a different industry solve a structurally similar problem. They figured the answer to housing might not be a bigger factory. It might be a smaller, smarter, more numerous one.

Three people who once deployed half a million robots

Vikas Enti, Felipe Polido, and Aaron Small met inside Amazon Robotics, where between them they helped put more than 500,000 robots to work across the world's most demanding fulfillment network. They learned, in fine detail, how to make automation flexible, cheap, and reliable at scale - the exact traits construction robotics had always lacked.

In 2022 they left to start Reframe. Enti, the CEO and an MIT graduate, has been candid that the motivation was not purely technical. He has said he wanted to do meaningful work on climate and, in his words, to make his kids proud. Polido took the CTO seat over robotics and software; Small, as COO, took on scaling the messy physical operation. They joined Greentown Labs in Boston as a nine-person team and started building.

The bet was specific. If a vision-guided robotic workcell could be built for around $300,000 and still match a conventional $5 million machine, then the whole economic logic of offsite construction flips. Suddenly you do not need one cathedral-sized plant. You can afford many small ones, each tuned to a local market, each adapting to local code.

Caption: four years, one factory, and a surprising number of triple-deckers later.

Pixels to parts, then parts to porch

Reframe is vertically integrated - it designs and builds, acting as manufacturer and builder at once. The whole thing runs on software the company calls "pixels to parts." A home design goes in; a set of manufacturable parts, already adapted to a municipality's zoning and safety code, comes out. That is the trick that makes "mass customization" possible: high volume without a cookie-cutter result.

The numbers it reports are the argument. A single microfactory targets five single-family homes a week. End to end, Reframe says it can deploy in about 100 days and cut construction cost by roughly 35%. The industry standard for assembling a module is measured in days; Reframe's stated ambition is to measure it in hours.

Caption: the cheapest part of the operation is the robot. The expensive part, as always, is the permit.

Money, modules, and a few finished roofs

In August 2025, Reframe closed a $20 million Series A co-led by Eclipse and VoLo Earth Ventures, with MassMutual Catalyst Fund, Cubit Capital, Planetary Health at RA Capital Management, Saga Ventures, and Nor'easter Ventures alongside. Twenty million dollars does not rebuild American housing. It does buy enough runway to prove the replication model works more than once.

The investor list is the tell. A robotics fund, a climate fund, and an insurer all looking at the same company suggests Reframe sits at an unusual intersection - automation, decarbonization, and the very practical business of putting roofs over people. In 2026 it added a top US housing innovation award and an MIT News feature to the pile.

The factory you can copy is the one that scales

A million homes by 2045 is the kind of number that invites an eye-roll, and Reframe knows it. The interesting part is not the figure - it is the method. Big factories scale by getting bigger, which makes them slower to build and riskier to bet on. Reframe is trying to scale the way software does: get one unit right, then duplicate it. Each new microfactory is a copy of a known thing, dropped into a region that needs homes, tuned by software to that region's code.

There is an obvious irony in a company full of warehouse roboticists deciding that the hardest logistics problem in America is the single-family house. But the wildfire rebuilds make the case plainly. When a neighborhood in Altadena burns down, the need is not for a clever home of the future. It is for a real home, fire-adapted, soon, at a price a displaced family can carry. A network of nearby microfactories is, at minimum, a more honest answer than a distant mega-plant.

Back in Andover, the robot finishes its wall and the person beside it takes over. Four years ago this room was an idea pitched by three people who had left a very good job to do something harder. Now it is a working floor that turns designs into airtight, all-electric homes, and the company's whole plan rests on the unglamorous promise that it can do this again somewhere else, and then again. The first microfactory was the experiment. The next one is the entire point.