SR&ED Case Study: Computer Vision Under Real-World Conditions
Background A manufacturer’s quality-inspection system used a computer vision model trained and validated under controlled lighting on a single product line. Expanding to a second production line introduced variable lighting, reflective surfaces, and a wider range of defect types. The Challenge The existing model’s accuracy dropped sharply under the new conditions, and it was unclear […]
SR&ED Case Study: Generative AI Development
Background A software company wanted to use a generative model to produce technical documentation drafts from structured engineering data, but standard fine-tuning of an off-the-shelf language model produced outputs that were fluent but frequently factually inconsistent with the source data. The Challenge Existing fine-tuning and prompt-engineering techniques could not reliably constrain the model’s output to […]
SR&ED Case Study: AI Fraud Detection
Background A payments company’s existing rules-based fraud system was missing new fraud patterns while generating too many false declines on legitimate transactions. Fraudulent transactions represented less than 0.1% of volume, and patterns shifted every time fraudsters adapted to detection. The Challenge The team needed a model that could detect rare, evolving fraud patterns in near […]
Can AI Development Qualify for SR&ED?
AI and machine learning development is one of the most active areas of SR&ED claims in Canada — and one of the areas the CRA scrutinizes most closely, particularly since 2026 guidance sharpened the line between eligible model development and routine application of existing tools. What Typically Qualifies What Typically Doesn’t Qualify Where 2026 CRA […]
SR&ED Case Study: When Laboratory Research Becomes Experimental Development
Background A contract testing laboratory ran standard analytical tests for industrial clients using validated, published methods. A client asked whether the lab could test a material class outside the range any existing validated method was designed to cover. The Challenge No accredited method existed for the material at the required sensitivity and matrix conditions. The […]
What Documentation Do You Need for an SR&ED Claim?
The CRA has been explicit about this: documentation is the single most important factor in whether an SR&ED claim is accepted. Good documentation doesn’t need to be extensive or created for the purpose of the claim — it needs to be contemporaneous, meaning created as the work happened rather than reconstructed at tax time. Core […]
The Difference Between Routine Engineering and SR&ED
Engineering teams are often surprised to learn that not everything they build qualifies for SR&ED — and just as often surprised to learn that work they assumed was “just engineering” actually does. The line comes down to whether the work applied known methods to a known problem, or generated new technical knowledge to resolve a […]
What Counts as Technological Uncertainty Under SR&ED?
Technological uncertainty is the starting point of every SR&ED claim. Get this concept wrong and the rest of the claim — systematic investigation, technological advancement — doesn’t hold together. Get it right, and it becomes much easier to see which of your projects actually qualify. The CRA Definition According to CRA guidance, technological uncertainty exists […]
Common SR&ED Eligibility Misconceptions
Every year, Canadian businesses that are clearly doing eligible R&D never file an SR&ED claim — usually because of a handful of persistent misconceptions about what the program actually requires. Here are five of the most common. Myth 1: “We’re not a tech company, so we don’t qualify.” SR&ED is defined by the nature of […]
Does Your Business Qualify for SR&ED? 10 Questions to Ask
Most businesses that qualify for SR&ED never apply — not because the work doesn’t count, but because no one asked the right questions. Run your current or recent projects through these ten questions to get a fast read on whether an SR&ED eligibility assessment is worth pursuing. 1. Did you try to solve a problem […]