SR&ED for Software & SaaS
If your engineering team is solving problems that don't have known solutions, SR&ED tax credits can fund a significant portion of those salaries — and now, AI compute infrastructure costs too.
Check If Your Work QualifiesWhat Qualifies in Software & SaaS?
- Algorithm development with technical uncertainty — where no known solution or library exists
- AI/ML model training and architecture: novel approaches to model design, training efficiency, or accuracy targets not achievable with off-the-shelf frameworks
- Custom compiler or interpreter development
- Novel database architecture and storage systems
- Performance optimization beyond what is achievable with known techniques or tooling
- Developing new software tools or frameworks where existing solutions are inadequate for the technical problem
How Bill C-15 (2026) Affects Software & SaaS
The doubled expenditure limit to $6M means software companies with larger R&D teams can now claim significantly more. GPU and compute infrastructure costs are now eligible as capital expenditures for the first time since 2014 — a major win for AI/ML companies investing in training hardware.
Why a Former CRA SR&ED Claims Reviewer Makes the Difference
Kazem Naderi spent years at the CRA reviewing and assessing SR&ED claims across industries. He knows exactly what CRA looks for, what documentation survives scrutiny, and what language gets claims approved. That insider perspective is something no other consultant can offer.
Book a Free Eligibility Call