EN
Back to the archive

The archive · AI & Models · Product decision · 2019–2024

Tractian's industrial-AI bet: $120M Series C, Forbes AI 50, ~R$250M revenue

Brazilian-founded industrial AI startup bets factories subscribe to cheap sensors plus AI predictive maintenance; $120M Series C in Dec 2024

Tractian

The betThat mid-size factories would buy cheap sensors plus AI that predicts breakdowns, sold as a subscription instead of an enterprise deployment.Scaling

What the business is

A manufacturing-AI company selling wireless sensors that monitor motors, pumps and compressors, with software that turns vibration and temperature data into predictive-maintenance alerts, diagnostics and work orders.

Starting capitalR$1.05B (~US$200M) across four rounds per Bloomberg Línea; latest US$120M Series C (Dec 2024) led by Sapphire Ventures with General Catalyst, Next47 and NGP Capital; earlier US$15M Series A led by Next47 with Y Combinator

How it started

Founded in Brazil in 2019 by Y Combinator alumni Igor Marinelli and Gabriel Lameirinhas, who had worked as software engineers at International Paper and seen how backward industrial machine monitoring was, together with Leonardo Vieira; Marinelli's father worked on factory floors, which kept the founders anchored to the maintenance problem.

What happened

Raised a US$15M Series A led by Next47 (May 2022) with Y Combinator, moved its HQ to Atlanta with offices in São Paulo and Mexico City; by 2024 it was the only manufacturing-focused company on the Forbes AI 50, had filed 12 patents, and grew to ~R$250M annual revenue with 40% from international markets; closed US$120M Series C in Dec 2024 led by Sapphire Ventures, valuing the company at ~R$4B (~US$720M).

How it ended up

Still scaling: Series C funds go to a new Brazil laboratory, R&D hiring and acquisitions, with customers like Suzano, Dexco, M. Dias Branco and Nissan Brasil (as of late 2024).

Background

Tractian was founded in Brazil in 2019 by Igor Marinelli, Gabriel Lameirinhas and Leonardo Vieira. Marinelli and Lameirinhas had been software engineers at International Paper, where they saw how backwards industrial machine monitoring was; Marinelli's father worked on factory floors, keeping the founders close to the maintenance problem. The bet: factories of almost any size would adopt a cheap sensor plus AI subscription that predicts breakdowns before they happen.

The product attaches wireless sensors to rotating assets — motors, pumps, compressors — and transmits vibration, temperature and electrical data over cellular networks. Software turns that data into diagnostics, alerts, checklists and work orders, and the models learn each machine's specific failure signature over time. TechCrunch covered the US$15M Series A led by Next47 in May 2022, reporting roughly 200 customers including John Deere, Bosch, Embraer and Hyundai.

By 2024 Tractian had moved its headquarters to Atlanta with offices in São Paulo and Mexico City, was the only manufacturing-focused company on the Forbes AI 50 list, and had filed 12 patents. Bloomberg Línea reported ~R$250M in annual revenue, 40% of it from outside Brazil, with clients such as Suzano, Dexco, M. Dias Branco and Nissan Brasil. In December 2024 the company closed a US$120M Series C led by Sapphire Ventures, with General Catalyst, Next47 and NGP Capital, at a ~R$4B (~US$720M) valuation.

Tractian's approach to the classic predictive-maintenance category was to make it cheap and self-service: rather than selling big enterprise projects, it sells subscriptions that include the hardware, letting maintenance teams start with one motor and expand from there.

What has to be true

  • Unplanned downtime is a huge, measurable cost — Siemens research cited by Tractian puts it at 11% of annual revenue for the world's 500 largest companies — so the pain is real and universal.
  • Selling hardware-inclusive subscriptions removed the upfront cost and integration risk that kept predictive maintenance out of small and mid-size factories.
  • Brazilian founders brought an emerging-market mindset: build something affordable for factories that cannot hire reliability engineers, then expand to the US and Mexico.
  • A consumer-grade deployment experience (sensor on, alerts on your phone) beat incumbent enterprise software on speed to value.

What can be applied

Predictive maintenance wins as a subscription, not a project: cheap sensors plus AI that learns each machine's failure signature let smaller factories get enterprise-grade uptime.

Aftermath

As of end-2024 Tractian was scaling on the back of its US$120M Series C: it planned a new R&D lab in Brazil, continued patent filings and engineering hiring, and pursued acquisitions, with annual revenue around R$250M and 40% of it generated outside Brazil across São Paulo, Mexico City and Atlanta operations.

Sources

spotted an error? The archive wants to know.

Your turn

You just read one. Describe what you are building, and see who is betting on the same thing.

Free account · 3 free questions · no card

Related cases