AI Is Not the Product. It's the Engine.

Lately, it seems like every VC firm has “pivoted to AI.” We have not. Not because we do not believe in the power and promise of AI, quite the opposite in fact.  Rather, we believe AI is a tool that enhances the functionality and utility of SaaS and robotics products, not a product in and of itself.  

As powerful a tool as AI may be, in our experience, no product, however impressive, will ever be of much use (nor sell particularly well) to enterprise companies unless it comes off the shelf capable of solving customers’ most pressing problems (i.e., without further customization), and with a UI designed to allow non-technical end users to quickly grasp its capabilities and easily incorporate that technology into their daily workflows.

We’ve always backed enterprise SaaS and robotics companies solving practical, high-friction problems in industries who, understanding that digital transformation is paramount to remaining competitive, are welcoming of new technology, but also require solutions that can be seamlessly employed, with as close to zero customization or employee retraining as possible.  

In other words, when you’re investing in practical, utilitarian tools for the dull, dirty, and dangerous jobs, AI doesn’t show up as a shiny overlay. It sits just below the surface, providing extraordinary functionality and exceptional customer ROI.

Quietly. Powerfully. Natively.

The truth is that AI, both predictive and generative, is now embedded across our entire portfolio. Not because we set out to find "AI companies," but because founders building real products today naturally reach for the best tools available.

And AI is increasingly one of them.

What We’re Actually Seeing in Our Portfolio

Some numbers that tell the story better than headlines:

• 60% of our post-ChatGPT investments use generative AI as a core product component

• 40% of those companies are building solutions that couldn’t have existed without genAI

• 55% of pre-ChatGPT portfolio companies have now added generative AI features

• 80% of companies added since 2023 are using predictive AI

• 100% of our post-2023 companies use either predictive or generative AI in a core function

AI isn’t a filter we apply after the fact. It’s increasingly a foundational element that helps make good software exceptional.

Our Approach: AI as Feature, Not Centerpiece

The best-performing AI-powered products in our portfolio all have a few things in common:

• They solve specific, high-stakes problems

• They fit into users’ existing workflows with minimal friction

• They make insights actionable for end users and visible to decision-makers

• They create measurable ROI, often by reducing time, waste, or risk in critical operations

It’s not just about the sophistication of the model; it’s about how that model delivers value to the user. AI on its own doesn’t sell. AI that helps a heavy-equipment reseller instantly diagnose maintenance issues or a utility company automate storm damage assessments?  That sells.

Why We’re Skeptical of the "AI Everything" Narrative

We don’t have a problem with ambition, but we do take issue with inflated claims. A few that we hear often, and don’t buy:

• "AI will replace most engineers within a year." GenAI coding tools are powerful. We’re seeing about a 3x improvement in developer efficiency across our companies. That’s a big deal.  But they’re not replacing human engineers any time soon, still making far too many mistakes and wrong turns to be left alone.

• "We’re on the cusp of AGI." Generative AI appears more human because it uses human language to communicate with users. That doesn’t mean it has anything close to human-level intelligence. After all, Predictive AI algorithms (those processing and learning from statistical data) have been in widespread use for at least two decades, and we haven’t heard anyone suggest that any of these platforms are showing signs of “general intelligence”.

• "Reaching AGI would be something to celebrate." Why? The best software platforms don’t replicate humans; they augment them by performing tasks we can’t, don’t want, or aren’t particularly well-suited to doing. Similarly, there are tasks that humans are better suited to than software is or likely ever will be.  Wouldn’t it make a lot more sense to focus on software’s unique strengths and leave the rest of human intelligence to the abundance of intelligent humans we already have?

AI doesn’t need to become something it’s not to be transformative. What it’s already doing, when applied with purpose, is more than enough.

Our Take

We aren’t chasing AI. We’re backing companies that are building better, faster, smarter tools for hard, complex problems plaguing essential but overlooked industries. Some of those tools happen to be powered by AI. Some of them wouldn’t work without it.  All of them seamlessly deliver immediate and comprehensive solutions to their users’ highest-leverage problems.

If that’s the kind of company you’re building, we should talk.

Read the full breakdown.