MaintainX

It is a Tuesday morning in a Cintas plant. Somewhere, a pump is about to fail.

It will not fail today. A vibration sensor strapped to its housing has been quietly streaming data for weeks, and a model trained on a few million machine-hours has already filed a work order on a technician's phone. By the end of the shift, the pump is fixed and nothing has shut down. That, in one ordinary scene, is what MaintainX is selling - and what eleven thousand industrial customers are now buying.

The company calls itself an AI-powered maintenance and asset management platform, which is a sober way of saying it has turned the paper clipboard into a $2.5 billion software business. Cintas runs on it. So do Duracell, Univar Solutions, McDonald's, Shell, Magna, Michaels and Titan America. Eleven million physical assets - mixers, conveyors, HVAC, generators, ovens, trucks - sit inside its database. Each one has a profile, a service history, and increasingly, a forecast.

A statistic, and a clipboard.

In 2018, Chris Turlica was an Entrepreneur in Residence at Deutsche Telekom Capital Partners, paid to think about mobile enterprise software. The job mostly involved reading research reports, which is not the kind of origin story that lands well at parties. Then he found a number that did.

Roughly 80% of the world's workforce does not sit at a desk. They drive trucks, run mills, mop hospital floors, fix conveyors, swap filters. And yet only about 1% of enterprise software spending ever reached them. The frontline economy was, in software terms, more or less unserved. The default tool for organizing the work that physically keeps the planet running was a clipboard, a binder, and - on a good day - an Excel file someone last touched in 2014.

FIG. 2 - The competitive landscape in 2018, as it appeared to a former CFA candidate with too much time on his hands.

The existing CMMS market - computerized maintenance management systems - was not exactly empty. It was just bad. The category had been dominated for two decades by enterprise software built in the IBM Maximo era: heavy, on-premise, license-gated, and designed primarily for the procurement officer who bought it rather than the technician who used it. Adoption rates were predictably grim. Frontline workers, asked to operate ten-year-old desktop apps from a phone they bought themselves, mostly just kept using the clipboard.

Four people, one phone-shaped wedge.

Turlica was joined by three co-founders - Hugo Dozois-Caouette as CTO, plus Mathieu Marengère-Gosselin and Nick Haase. Three of the four are Canadian; the company was incorporated in Delaware, headquartered in San Francisco, and built largely on Pacific and Eastern time zones at once. Turlica's previous startup, Voo, was a consumer app. He is, on paper, a CFA charterholder turned product person, which is the kind of CV that tends to read better in retrospect than it does at a Series A pitch.

The bet was simple, almost suspiciously so. Build the maintenance app that frontline workers would actually open. Mobile-first. Free at the bottom. Modeled less on Maximo and more on Slack - something a maintenance lead could roll out to a plant without asking IT's permission. Then, once you owned the workflow, layer in the things only software can do: photos, real-time chat, parts inventory, preventive schedules, and eventually, predictive AI fed by sensors on the machines themselves.

Bain Capital Ventures led the seed in 2020. Bessemer led the Series B. By December 2023 the company crossed a billion-dollar valuation with a $50M Series C, and in July 2025 it added a $150M Series D that pushed the valuation to $2.5 billion. Total raised: $253.8 million. The investor list now reads like a roll call: Bessemer, Bain Capital Ventures, D.E. Shaw, August Capital, Amity, Founders Circle, Sozo, Fifth Down. Even the angels - Rahul Mehta of DST Global, Dave McJannet of HashiCorp - are not subtle.

A work order, but with opinions.

What MaintainX actually does is unglamorous in the best possible way. A technician sees a problem - oil leak, weird noise, blinking light - and opens the app. They photograph it, attach a manual page, tag the asset, assign a teammate, and hit submit. A supervisor approves it. A part is reserved. A preventive maintenance schedule shifts. When the job is done, the asset's record updates, and so does the plant's compliance log.

The trick is everything around the work order. Procedures are digital and templated, which means the new hire on shift three follows the same fifteen steps as the senior tech on shift one. EHS checklists are baked in. Parts and purchase orders sync to SAP, NetSuite, Oracle EBS or Microsoft Dynamics. SCADA alerts from Inductive Automation's Ignition platform create MaintainX work orders automatically, closing the loop between the OT side of the plant and the IT side of the business.

And then, increasingly, there is the AI layer. MaintainX CoPilot drafts procedures from a few lines of text. Anomaly detection watches vibration and temperature data for machines that are starting to misbehave. Predictive insights reorder the preventive maintenance schedule based on what each individual asset is actually doing, rather than what its manufacturer suggested in 1998.

The customers, with logos turned slightly sideways.

The least interesting way to prove a B2B company is real is to list its logos, which is also, awkwardly, the most effective. MaintainX's customer list is mostly Fortune-adjacent industrial companies that do not give out reference calls casually.

FIG. 3 - Companies that have agreed, in writing, to let MaintainX into their plants.

The numbers around the customers are arguably better than the customers themselves. G2 reviewers - users, not analysts - have ranked MaintainX at or near the top of the CMMS category since 2022, which is a much harder leaderboard to fake than an analyst quadrant. Average ratings sit above 4.6. Adoption rates inside customer accounts, the metric most CMMS vendors quietly do not publish, are reportedly the highest in the category.

Less downtime, fewer injuries, more boring Tuesdays.

It is tempting to describe maintenance software in heroic terms. The reality is more modest. Every percentage point of unplanned downtime in a manufacturing plant is, depending on the industry, somewhere between a six- and an eight-figure problem. Every safety incident is at least one person who went home worse than they arrived. MaintainX's pitch, stripped of marketing, is that better software makes both numbers smaller.

The company's stated mission - to create a better day for the world's frontline workers - sounds, on first read, like the kind of line a copywriter writes on the way out the door. On second read it is doing slightly more work. The product is built for the technician first, the supervisor second, and the procurement officer somewhere a polite distance behind. That ordering is unusual in enterprise software, and it is the single hardest thing to fake.

The interesting part is not the app. It is the data.

Eleven million assets, instrumented, logged and increasingly sensed, add up to something the CMMS category has never had before: an actual longitudinal dataset of how industrial machines behave in the wild. That is the bet sitting underneath the Series D. The work order is the wedge. The asset history is the moat. The AI is the multiplier.

The competition - UpKeep, Limble, Fiix (now Rockwell), IBM Maximo, eMaint, Brightly - is real, and the category is large enough that more than one of them will end up doing fine. MaintainX's advantage is less about features and more about the order of operations: get the technician to open the app, then earn the right to instrument the machine, then earn the right to predict its failure. Most of the category tried to do those steps in reverse, which is why the category has, until recently, been mostly invisible.

It is a Tuesday morning in a Cintas plant. The pump is humming. The clipboard is in a drawer somewhere. A technician taps a phone, marks the work complete, and walks to the next asset. Nothing shuts down. Nothing makes the news. Which is, when you think about it, the entire point.