Dave Killeen walked onto the SaaSiest stage wearing AI-powered glasses that streamed his speaker notes directly into his field of vision. It was a small detail, but it perfectly captured the bigger message behind his keynote.
We are entering an era where leverage matters more than effort.
And according to Dave, most SaaS companies are still thinking too small about what AI actually changes.
For years, the conversation around AI has been strangely comforting. We kept telling ourselves that while AI could generate code, summarize meetings, or automate repetitive work, it could never replace human taste. Humans would still be the chefs designing the menu while AI stayed in the kitchen doing prep work.
Dave used to believe that too.
Now he thinks that mindset is dangerously outdated.
The shift happened when AI crossed what former Tesla AI leader Andrei Karpathy called “the threshold of coherence.” Models stopped feeling like unreliable assistants and started behaving like dependable collaborators. Suddenly the bottleneck was no longer technical execution. It was clarity, coordination, and speed of learning.
And that changes everything.
Small Teams Are Now Competing Like Giants
At Pendo, where Dave works with enterprise SaaS companies on product strategy and AI adoption, he is seeing a completely different operating model emerge. Small teams are now scanning markets faster, building faster, and validating ideas faster than organizations ten times their size.
Not because they work harder.
Because they have built systems that compound.
Meet “Dex”: Dave’s AI Operating System
Dave shared how he built “Dex,” his own AI operating system that connects tools like Slack, Salesforce, Gong, email, and external market signals into a single intelligence layer. Every morning, the system reviews information against his goals, prioritizes what matters, and even calls him out when he is avoiding detail.
The important point was not the product itself.
It was the realization that AI is no longer just helping people execute work. It is helping people coordinate work, discover opportunities, and accelerate decision-making loops.
That is where SaaS leaders should pay attention.
The companies winning in the AI era will not necessarily be the ones with the biggest engineering teams or the largest data sets. They will be the ones who shorten the cycle between signal, execution, and customer validation.
The New AI Playbook for SaaS Leaders
Dave summarized this as a new three-step playbook:
1. Out-hunt
2. Bend the system
3. Prove impact
The first step, “out-hunt,” is about building systems that constantly scan the market for weak signals before they become obvious. Dave spends a huge amount of time looking at open source ecosystems like GitHub because many future product categories start there long before they show up in analyst reports or competitive battlecards.
But no human can manually keep up with that pace anymore.
So instead of personally monitoring every trend, he built AI agents that scan repositories, summarize developments, score ideas, and even generate recommendations automatically. In his words, product leaders can no longer afford to act only as “head chefs.” They need to become “truffle hunters” who build systems that discover opportunities at scale.
The second step, “bend the system,” might be the most important insight of the entire talk.
Most organizations are treating AI like a productivity tool. Faster emails. Faster code. Faster presentations.
Dave argues that this completely misses the point.
He referenced the shipping container revolution in global trade. Containers did not simply make shipping cheaper. They transformed the entire structure of supply chains, ports, labor, and globalization itself.
AI is doing the same thing to knowledge work.
The real opportunity is not simply accelerating existing workflows. It is redesigning how work itself gets coordinated.
Dave showed how he built a completely custom AI-native product workflow where moving ideas across a Kanban board automatically triggers different AI agents. One agent critiques ideas harshly. Another transforms them into detailed PRDs. Others generate production-ready code.
In one example, the system generated over 17,000 lines of code in roughly two and a half hours.
Whether that exact workflow becomes mainstream is almost beside the point. The bigger takeaway is that SaaS companies now compete against organizations that are redesigning how software gets built from the ground up.
And finally comes the third step: prove impact.
When building becomes dramatically cheaper, shipping features is no longer the achievement. Customer evidence becomes the differentiator.
This is where Dave connected the conversation back to Pendo and product analytics. In a world where AI can generate endless ideas and prototypes, the real advantage comes from understanding what customers actually use, where friction exists, and whether a product change creates measurable value.
That feedback loop becomes the moat.
The Companies That Learn Fastest Will Win
The strongest companies in the next generation of SaaS will not be the ones building the most features. They will be the ones learning the fastest.
What made Dave’s keynote resonate was not the technology itself. It was the mindset shift underneath it..
The AI era is not just another tooling upgrade for SaaS.
It is a complete rethink of how modern software organizations discover opportunities, coordinate work, and create value.
Or as Dave put it:
“When everyone can build faster, the winners will be the teams that know what to build, why it matters, and whether it actually works for customers.”
