Linear #139: Will vSaaS Become Fast Fashion? Tractian: Vertical AI + Hardware for Industrial Maintenance

One vSaaS breakdown. One biz story. One 'how to'. In your inbox once a week.

---

Today’s newsletter is sponsored by Unit, the leading embedded finance platform for vertical SaaS.

Become the financial OS your customers rely on. With Unit’s Ready-to-Launch solutions, you can embed capital, banking, and bill pay in weeks - zero build required. Trusted by platforms like Bill.com, HoneyBook, and Homebase.

Learn more at unit.co

---

Alright, let’s get to it…

Will SaaS Become “Fast Fashion”?

There is a lot of fear circling about Sam Altman’s recent tweet, comparing SaaS to the “fast fashion industry.” He’s basically saying that SaaS will go the way of flashy / fast trends, with little to no staying power.

While I understand that helps Sam’s fundraising case, IE saying that they are going to own the infrastructure layer AND the app layer, I don’t see it happening…

Another interesting take I hard on this, was from Chris Paik, who basically believes “wrappers” or simply Apps that sit on top of LLM’s / infrastructure, only really work if they are on top of a commodity.

To which I followed up with:

So there a few schools of thoughts with what happens next that I’m starting to see emerge. You have to pick one that you believe and bet your time/capital there.

1. Infrastructure/LLM companies are able to not only own the underlying technology but also own the App Layer

2. Infrastructure/LLM companies compete like hell with each other to continue to perfect their models and a bunch of App Layer companies emerge and build massive companies on top of their infrastructure tech

3. Some hybrid of the above two

I think a good parallel here is Steve Jobs early belief when he was working on the iPhone. Steve famously DID NOT wan’t to allow any third-party apps inside the iPhone. He wanted EVERY APP to be made by Apple. He felt quality would be horrendous if they allowed anyone to create an app.

With the history in hindsight here, he obviously changed his mind, and hundreds of massive companies were created on top of his monopoly. Did he take his bite of flesh from all of these? Yes. But there is just no way Apple, even as one of the worlds biggest company, could undertake that. If they did, would they have the staying power with the infrastructure? Probably not.

I also look at Google’s search monopoly as an interesting example. They have obviously moved to build App Layer technology in certain verticals, Education / Healthcare / etc all comes to mind. Some of these worked, some didn’t, but none of them ever achieved monopoly status, and their main focus has always remained on fighting off folks coming for their money-maker / cash cow. This, arguably, is what has caused them to miss the early boat of the AI revolution.

So my view is that AI/LLM companies look the same as infrastructure companies have always looked in every new technology era. They will take their bite of flesh from App Layer companies but they are never going to own the complete App Layer. Maybe they own a few verticals, or become significant players in some, but their focus will constantly be on defending their castle — there is hyper-competition at the infrastructure level right now and probably always will be.

While building technology will continue to get easier and easier, the areas to capitalize are making very industry-specific technology that sits on top of AI infrastructure.

Tech will become a commodity,

Domain expertise (workflows, apps solely trained on industry data, industry-specific GTM motions, etc.) is and will continue to be a moat.

I’d love to hear your thoughts on this piece. Where you think I’m right/wrong. Reply to this email and I’m happy to pull them together for a future newsletter piece so we can all continue to learn together.

The tides are changing, let’s stay ahead of it, together !

Basically Last Call On
The vSaaS Summit!

We’re almost sold out. Grab your ticket! Prices go way up September 1st. If you’ve been putting off buying your ticket, now is the time to pull the trigger.

Going to be a legendary time. Here’s a few incredible folks who are speaking:

• Founder/CEO of Slice ($1B+ EV)

• CRO of Owner.com ($1B+ EV)

• CEO of PatientNow ($1B+ EV)

• Founder & CEO of Valsoft ($1B+)

• Former Chief Growth Officer of Service Titan ($1B+)

• Co-Founder & CEO of Unit

• Founder & GP of Euclid VC

• Founder & GP of TipTop VC

• GP of Frontier Growth

• Managing Partner of Reformation Partners

• Founder & CEO of Quadshift

• Managing Partner of Blueprint equity

• Founder & CEO of Bungalow

• Founder & CEO of Roofer.com

• Founder & CEO of Cents

• Partner at Google Ventures

• Co-Founders/Co-CEO’s of EvolutionIQ

• Founder & CEO of Tight.com

---

Tractian: Vertical AI for Industrial Maintenance

A deep dive into the Brazilian startup that's revolutionizing industrial maintenance with 340M+ daily ML inferences and a path to IPO

Most SaaS founders chase sexy problems. Igor Marinelli, Gabriel Lameirinhas, and Leonardo Vieira had a different obsession: solving the $50 billion annual cost of unplanned industrial downtime.

The founding trio shared something rare in tech: authentic domain expertise. CEO Igor Marinelli, son of Brazilian and Italian migrants, grew up watching his father struggle as a maintenance technician. By age 13, Igor was already programming inventory management systems for a Portuguese motorcycle manufacturer. His co-founders Gabriel and Leonardo came from similar industrial backgrounds, giving them credibility that Silicon Valley outsiders typically lack.

Founded in 2019, Tractian emerged from a simple but profound insight:
Industrial maintenance was still trapped in the stone age of reactive repairs.

When a critical machine fails in a manufacturing plant, entire production lines can halt, costing companies hundreds of thousands per hour. Yet most factories were still operating on "fix it when it breaks" mentalities.

The problem wasn't just technological, it was cultural and economic. Legacy industrial suppliers had been selling the same rudimentary monitoring solutions at exorbitant prices for decades, with multi-year implementation cycles that made procurement departments wince.

———————— Product Evolution————————

Phase 1: Hardware-First Approach (2019-2021)

Unlike software-only competitors, Tractian started with purpose-built hardware:

• SmartTrac sensors: High-frequency vibration and temperature monitoring

• EnergyTrac sensors: Power consumption and electrical fault detection

• Battery-powered design: Extended field deployment without infrastructure changes

The hardware strategy was deliberate, industrial customers trusted physical devices over cloud-only solutions. Installation took minutes, not months.

Phase 2: AI-Powered Software Platform (2021-2023)

• Intuitive dashboarding with real-time alerts

• Machine learning-based diagnostics and recommendations

• End-to-end CMMS capabilities for work orders, scheduling, and inventory

• Auto Diagnostics™ for automatic failure prediction

Phase 3: Industrial Copilot (2023-Present)

Today, Tractian positions itself as the "Industrial Copilot"—an AI-native platform that doesn't just monitor equipment but actively coaches maintenance teams. The technology stack includes:

• Three-layered AI system: Self-diagnostics, comparative analysis, and predictive modeling

• 340+ million daily ML inferences processing data from 100,000+ active sensors

• 11.5 million samples processed daily (130 per second)

• Native integration of sensors, AI, and CMMS in a single platform

———————— GTM Motion————————

Interestingly, they did not start going after the U.S. market.

• Dedicated enterprise sales teams in Brazil, Mexico, and the US

• Solution-oriented approach focused on ROI demonstration

• Technical sales process with on-site equipment analysis

• Fast deployment as competitive differentiation

Content-Led Demand Generation

• Educational content strategy: Case studies, whitepapers, technical blogs

• ROI-focused messaging: Up to 18x return on maintenance investment

• Customer success stories: Danone, Bosch, Hyundai, John Deere testimonials

• Industry event presence: Direct engagement at manufacturing trade shows

Geographic Expansion Strategy

• Brazil first: Dominate home market before international expansion

• Mexico second: Similar industrial base, cultural affinity

• US third: Largest market opportunity with Atlanta headquarters

• 50% sales growth in Mexico during 2024 nearshoring boom

The strategy worked: Tractian grew from 66 customers in 2021 to 500+ industrial logos today, boasting 150% net revenue retention,.

———————— How Big Can This Get?————————

Several metrics suggest Tractian is building towards $100M+ of ARR this year. The Industrial IoT market is massive, $300B+, so there is plenty of ceiling left here…

• Revenue growth: 5x growth in 2 years puts them on track for $100M+ ARR

• Market expansion: Multi-geography presence reduces single-market risk

• Enterprise customers: Fortune 500 logos provide IPO-quality reference base

• Technical scale: 340M daily inferences demonstrate operational leverage

• Funding runway: $120M Series C provides 3-4 years of growth capital

———————— Funding Journey————————

Tractian stepped up financing rounds pretty quickly in classic YC fashion, but most notably, they secured a $120M Series C in December of 2025 led by Sapphire Ventures and General Catalyst.

———Takeaways for Founders/Operators/Investors———

1. Domain Expertise Creates Unassailable Moats

All three founders came from industrial backgrounds. This wasn't academic curiosity, they lived the pain of broken maintenance processes. Their credibility with industrial customers stems from authentic understanding of operational challenges.

2. Hardware + Software = Higher Switching Costs

Pure software is a lot easier to displace than software + sensors/light hardware. Hardware integration creates stickier customer relationships and higher barriers to competition. Tractian's vertical integration enables competitive differentiation that's difficult to replicate.

3. AI as Core Product, Not Feature

Tractian didn't add AI to existing software, they built AI-native architecture from day one. This architectural decision enables competitive differentiation through 340M+ daily ML inferences that competitors cannot easily match.

4. Metrics Tell the Real Story

340M daily inferences, 150% NRR, and $32.2M revenue aren't vanity metrics, they represent real industrial transformation at scale. The technical metrics validate the business metrics.

Conclusion:

Tractian represents a rare combination of deep domain expertise, technical sophistication, and rapid commercial scaling. With $196M raised, 340M+ daily ML inferences, and a clear path to $100M+ ARR, they're building the definitive platform for AI-powered industrial maintenance. Excited to see their journey continue!

---

---

Have a product or service that would be great for our audience of vertical SaaS founders/operators/investors? Reply to this email or shoot us a note at ls@lukesophinos.com