The Digital Uplift
students, agentic engineering, local businesses
An experiment I pulled together with fellow students from North Island College, Vancouver Island University and the University of Victoria: how much could we help local businesses with agentic engineering while finishing our degrees, without making them burn their pockets. Money was never the goal.
By the end of 2024, AI had made a real, working site far cheaper to build. Vancouver Island is full of small businesses with no website, or one that loads slowly and ranks nowhere. So we built workflows that turned out small, clean sites, complete where it mattered, at the lowest price we could still deliver properly. Nobody was meant to lose out.
I took the build. Every question a local owner might search got its own static route, each with its own metadata and canonical, all prerendered so search engines get finished HTML.
Next.js with static generation, Shadcn for UI, next/image and self hosted fonts gave 100 Lighthouse scores in SEO, Accessibility and Best Practices. Every route led to the same booking path, and a Botpress webchat assistant answered questions after hours, my first production agent.
Finding clients was harder than building for them, and none of us could run it full time. When I joined Botpress in September 2025 we stopped taking new clients instead of running it at half attention. Existing work was finished, the site archived, and a notice added across it.
Distribution is the hard part. A perfect score opens a conversation but does not close one. Owning spec, design, code and support at once teaches scope discipline, and closing something cleanly is part of building it.