Hello to Canada’s SaaS and AI Community,
How much of the R&D your team is already doing could qualify for funding?
For Canadian SaaS founders, the answer might be more than they think.
Speaking with SAAS NORTH AI, Jude Brown, Co-Founder of SRED.ca, challenged common misconceptions about Canada’s Scientific Research and Experimental Development (SR&ED) program, explained where AI-assisted development fits in and shared a simple habit that could help founders uncover overlooked opportunities.
Key takeaways:
- Improving existing technology may qualify for SR&ED.
- AI-assisted coding doesn’t automatically disqualify R&D.
- Document technical challenges, experiments and failures quarterly.
- Explore non-dilutive funding early to help extend runway.
Co-Founder/Producer, SAAS NORTH Conference Editor, SAAS NORTH NOW
You Don’t Have To Invent Something New
Jude regularly meets founders who assume SR&ED is reserved for companies inventing technology the world has never seen.
It’s an understandable misconception. The Canada Revenue Agency (CRA) has to administer a program covering industries with very different kinds of research and development, which makes its guidance difficult to interpret for any one business.
But as Jude explains, innovation doesn’t always involve reinventing physics.
A SaaS company might be improving an existing product, trying to solve a performance limitation or developing technology to compete in an established market. A competitor may have solved a similar challenge already but that doesn’t mean its technical approach is available to everyone else.
After all, competitors rarely hand over their blueprints.
What matters for SR&ED is whether the company encountered technological uncertainty and undertook systematic investigation or experimentation to resolve it. Routine development doesn’t qualify simply because it required engineering effort.
Jude says the biggest misunderstanding is assuming that only entirely novel inventions count.
“A lot of the work you’ve already done qualifies. We just need to dig in and see if it meets the program criteria.”
For founders, it’s worth examining the technical work behind their products before dismissing a potential claim.
Does AI-Written Code Still Count?
AI coding tools have introduced another question for SaaS founders: if an agent writes the code, can the work still qualify for SR&ED?
Jude’s answer comes with an analogy.
“It would be akin to the CRA asking you what pencil you use to write the story.”
The tool isn’t the deciding factor. The technological uncertainty and the work undertaken to resolve it are what matter.
He uses Claude to illustrate the distinction. Imagine a developer gives the agent a problem, defines the requirements and receives a working solution three hours later. If there was no technological uncertainty to resolve, that activity alone wouldn’t establish an eligible claim.
Now imagine the agent proposes several approaches, each with limitations. The team tests them, encounters problems, changes direction and continues experimenting before finding a solution.
That process may contain qualifying R&D, provided it meets the program’s requirements.
“What actually matters ultimately is your documentation, the iteration,” Jude explains.
It’s an important distinction for AI-native companies. The speed at which code is produced doesn’t tell the full story of the development work. The technical questions, experiments and decisions made along the way are what founders need to understand and record.
Keep A Diary Of What Didn’t Work
Jude’s most practical advice is to start keeping an R&D diary.
Every three months, sit down with the technical team and ask what they’ve been working on. Which problems took longer than expected? Where did they get stuck? What approaches did they try?
He deliberately recommends avoiding the question, What SR&ED work did you complete? Developers may not recognize which activities could qualify.
“When you find failure, that’s when usually you find the pockets of SR&ED.”
He’s referring to the technical challenges that lead teams to experiment, abandon an approach or develop a different solution. Failure alone doesn’t establish eligibility, but the work behind it may provide useful evidence.
The difficulty comes when companies leave those conversations until tax season.
Jude describes the challenge of asking developers to reconstruct projects from more than a year ago. By then, the details have faded, team members may have moved on and the reasoning behind decisions can be difficult to recover.
His alternative is refreshingly simple. Record quarterly conversations, use AI to prompt developers about recent work and connect the notes to development tickets, issue trackers and meeting records.
A diary won’t replace the technical and financial evidence required for a claim, but it gives founders a much stronger starting point.
It also makes the process less stressful.
As Jude puts it, regular documentation “lowers the temperature in the room.”
Don’t Leave Funding Until Year-End
Jude believes founders should think about SR&ED and other non-dilutive funding opportunities long before they need the money.
For SaaS companies balancing product development, hiring and growth, programs such as SR&ED and the National Research Council’s Industrial Research Assistance Program (IRAP) may provide support without requiring founders to give up additional equity.
That matters in a funding environment where investors are paying close attention to how efficiently companies use their capital.
“Capital efficiency is definitely what investors are looking for,” Jude says.
But he also warns against assuming anticipated funding will arrive exactly as planned.
A founder might identify 12 development projects, complete only four and discover that those four don’t meet the eligibility criteria. If the business has already factored the expected refund into its cash flow, the shortfall can become a serious problem.
Understanding potential eligibility early makes financial planning more realistic.
Jude also encourages founders to consider whether they need outside help preparing a claim. Consultants can provide valuable expertise, but their fees should make sense relative to the work involved and the potential benefit.
His advice is particularly direct for early-stage companies:
“You do not have to pay a consultant. You can do it yourself, but you absolutely have to do it.”
The broader point is that founders need to understand their funding options and take responsibility for the process, whether they manage it internally or bring in professional support.
Make The Work Count
Jude’s advice starts with something most SaaS companies already have: a backlog of technical problems, experiments and decisions that shaped their products.
Some of that work may be eligible for SR&ED. Much of it won’t be. The important thing is knowing enough about what happened to make an informed assessment.
A quarterly conversation with the development team can help uncover opportunities that might otherwise be overlooked, while giving founders a clearer picture of the technical work their business is funding.
For companies already investing heavily in software development, that’s a habit worth building before another year of work disappears into completed tickets and forgotten conversations.