Building technology is getting faster. Proving it works, getting it to customers and turning it into a business are still the hard part.
The Bennett Jones executive breakfast at DEPLOY, the defence summit at SaaS North
I spent November 3rd to 5th at SaaS North in Ottawa, connecting with founders, investors, technology leaders, and organizations supporting Canadian startups.
The conference theme was Scaling in the Era of AI, with a new addition this year: DEPLOY, a summit focused on Canadian defence and dual-use technology.
There was plenty of discussion about how AI is changing software development, business models, and the way companies scale.
But several conversations stood out to me for a different reason.
Building technology is getting faster. Turning it into something customers trust, businesses can sustain, and markets will actually buy remains much harder.
Here are the themes I brought back to Waterloo Region.
3 daysat SaaS North in Ottawa, November 3 to 5
4 themesbrought back to Waterloo Region
1. AI makes building faster. It doesn't automatically make products better.
The discussion focused on the visibility needed to understand what AI agents are doing, particularly when they perform tasks independently.
As AI generates more code and handles more workflows, understanding what happened when something goes wrong becomes increasingly important.
Logging and tracing can help explain an agent's actions.
“
Knowing what an agent did is not necessarily the same as knowing whether it did the right thing.
That raises important questions for founders. How do you know an AI generated feature actually solves the customer's problem? Who is accountable when an agent makes the wrong decision? How do you catch failures before customers experience them?
After decades working with software companies, this was one of my strongest takeaways. Faster development does not eliminate the need for testing, communication, clear requirements, and shared understanding. In some circumstances, it makes those things more important.
2. Sovereign AI is about control, not isolation.
At the RBCx investor breakfast and a TELUS/L-SPARK session, sovereign AI was a major topic.
The conversations went beyond where servers are physically located. They raised questions about who controls infrastructure, what happens when a critical supplier becomes unavailable, how jurisdiction affects access to data, and whether Canadian organizations have meaningful alternatives.
The main point I gathered here was that Canadian sovereignty does not mean building every technology component domestically.
“
It means understanding dependencies and retaining enough control to make independent decisions.
Building innovative technology is one challenge. Getting access to enterprise customers, demonstrating a credible business case, and integrating into established organizations are others.
For Waterloo startups developing AI products, these are practical business considerations, not just government policy discussions.
Who owns the critical technology? Where does the data go? What happens if an infrastructure provider changes its terms? And can the business demonstrate that its technology delivers sufficient value to justify adoption?
3. Defence technology presents opportunities, but commercialization is still the hard part.
One of the more interesting additions to SaaS North was DEPLOY, which brought defence leaders, government representatives, investors, and technology companies into the same conversations.
At the Bennett Jones executive breakfast, the central question was how Canada can become a better customer for Canadian defence innovation.
A particularly relevant perspective came from Mike Peasgood, CEO of Waterloo based ENVGO, which develops autonomous and advanced marine technology. Mike previously co-founded Aeryon Labs, another Waterloo success story.
He walked through the path from a prototype to a buyer, and the step that is familiar to many founders.
First
Get the technology into the hands of actual operators
Then
Test it in real world conditions
Then
Improve it rapidly based on their feedback
The hard part
Get from a promising demonstration to a real procurement contract
Government funding and early testing opportunities are useful, but companies also need predictable demand, purchasing timelines, and customers willing to adopt their technology.
For a Canadian startup trying to compete internationally, having its own government as an early customer can provide credibility far beyond the value of the initial contract.
The opportunity is not limited to companies building military hardware. AI, cybersecurity, autonomy, software, sensing, and other technologies can have civilian and defence applications.
But founders need to understand the actual operational problem and the requirements of the market before deciding whether defence is a viable direction.
4. The connections between companies may matter as much as the technology.
One of the most valuable parts of SaaS North was the opportunity to speak directly with founders, investors, accelerators, and people working to commercialize Canadian technology.
Organizations such as L-SPARK and Invest Ottawa are part of that broader picture, connecting companies to expertise, enterprise relationships, programs, and potential customers.
Across those conversations, I kept returning to a familiar issue: companies can be very good at building their products while struggling with the relationships and organizational capabilities needed to grow.
Sometimes the missing piece is not another developer, another feature, or another funding application.
It might be an introduction to someone who understands a market, access to a real customer, independent expertise, or better communication between the people building the product and the people responsible for its success.
This is also where regional technology communities have an important role to play.
What does this mean for Waterloo Region?
Waterloo already has a history of producing companies that compete internationally. ENVGO and Aeryon Labs are useful examples of technology and expertise developed here finding applications in demanding global markets.
But strong engineering capabilities alone do not guarantee commercial success.
The conversations at SaaS North reinforced several questions worth asking when building a company:
These questions apply to a student building a first product, an early stage startup looking for its first customers, and a scaleup entering international markets.
My biggest takeaway
SaaS North was about how Canadian companies can turn technical capability into sustainable businesses, not just a conversation about artificial intelligence.
AI is changing what can be built and how quickly. Sovereign technology is raising new questions about control. Defence and dual-use markets are creating possibilities for companies that might not previously have considered them.
But behind all of that are the same fundamental challenges: understanding customers, building reliable products, finding buyers, developing relationships, and executing well.
For Waterloo's builders, I think the opportunity here is to get better at turning technology into something valuable, trusted, and commercially successful.
A Waterloo consulting firm that helps B2B SaaS leaders restore release confidence when scaling, organizational change, or AI driven software delivery creates new quality risks.
