Product platform
Finding the genuine few among the many.
BuyCanada.ai · AI product discovery platform · built end to end
A discovery platform that helps Canadians find products actually made in Canada. Vector search and a retrieval-based recommendation engine do the sorting, because origin claims are easy to make and hard to verify.
What was actually delivered.
Count them. This is the list that stops an engagement being priced as a single deliverable.
- 01Product design
- 02Information architecture
- 03Vector search implementation
- 04RAG recommendation engine
- 05Origin verification logic
- 06Front-end build
- 07Deployment and infrastructure
- 08Ongoing iteration
Keyword search could not answer the question
People do not search for a product name, they search for an intention. Matching that to real inventory needs semantic search rather than string matching, which is a different class of problem to a normal catalogue.
Verification is the product
Anyone can claim a product is locally made. The value sits in the layer that checks, which is where most of the engineering went.
Built alone, start to finish
Product thinking, architecture, the search layer, the interface and the deployment. This is the shape of engagement where there is no agency and no second vendor to hand anything to.