How I Built 8 Vertical AI Products From Edmonton, Canada - And What I Learned

By Tilak Raj, CEO & Founder - Brainfy AI March 2026 Tags: vertical AI, founder lessons, Canadian startup, Brainfy AI, SaaS building

People ask me all the time: how do you run 8 AI products at once from Edmonton? The honest answer is with a lot of systems, some failures I do not talk about enough, and a thesis I refined over hundreds of hours of building.

Why I Started Building Vertical AI

In 2023, I looked at the AI landscape and saw a gold rush of horizontal tools. Chatbots. Writing assistants. General-purpose automation platforms. Everyone was building nearly the same thing with a slightly different UI and pricing model.

I had a different instinct. I had worked closely enough with small businesses in traditional industries to know they did not need a smarter chatbot. They needed a system that understood their workflow, language, compliance obligations, and legacy stack.

> The opportunity was not in making AI more general. It was in making AI deeply specific.

Horizontal AI was already crowded and commoditizing. Vertical AI had real defensibility.

Product 1: AgriIntel - Where Everything Started

AgriIntel was the first product I built after committing to the vertical AI thesis. Grain and oilseed operations in Alberta were making expensive decisions using gut feel and spreadsheet stitching.

I interviewed 20 farmers and agribusiness operators before writing code. The pattern was clear: the data existed, but there was no intelligence layer connecting it into action.

AgriIntel took six weeks from zero to MVP. The hardest part was not model orchestration. It was learning crop cycles, input cost dynamics, and weather impact deeply enough to produce outputs that were genuinely useful.

Lesson one: in vertical AI, domain knowledge is not optional. It is the product.

Products 2-3: Brainfy AI and CovioIQ

After AgriIntel, I understood the template:

1. Validate pain with 10-15 operator interviews. 2. Identify 2-3 workflows with measurable upside. 3. Build an MVP in 30 days. 4. Iterate only on real usage data.

Brainfy AI came next: an AI implementation platform for businesses without technical teams. The barrier to adoption was not cost. It was setup complexity. Brainfy removes that complexity with prebuilt workflows.

CovioIQ followed. Canadian insurance is heavily regulated, paper-heavy, and still tied to legacy systems. CovioIQ focused on underwriting support, claims acceleration, and customer operations, with Canadian regulatory context built in.

Products 4-5: Navlyt and RealtorDesk

Navlyt is my most technically demanding build. Charter aviation operators face FAA, Transport Canada, EASA, and internal ops-spec burdens. Navlyt tracks compliance status, flags gaps before audits, and prepares documentation for certification maintenance.

RealtorDesk solved a different category of pain: lead management, follow-up, and listing ops in real estate. It uses AI for lead scoring, follow-up personalization, and workflow automation in a vertical-native CRM model.

Products 6-8: PetVitale, CanadaCompliance, IndigenousRising

PetVitale: proactive pet health support with symptom triage, wellness tracking, and nutrition planning.

CanadaCompliance.ai: AI-assisted Canadian compliance tracking and documentation across privacy and industry obligations.

IndigenousRising.ai: mission-driven tooling for Indigenous entrepreneurs, including grant discovery and startup support aligned with community governance realities.

What I Would Tell Someone Starting Today

Pick one industry and go deep

I built 8 products, but each is narrow at launch. One industry. One painful workflow. One user type.

Your moat is not the model

Models are commoditizing. Durable value comes from workflow integration, domain data, and trust inside a specific user community.

Validate before you build, every time

I still run 10-15 interviews before new product builds. Those interviews still change my assumptions almost every time.

Canada is a feature, not a limitation

Building in Edmonton gives access to agriculture, energy, and public-sector buyers that often require Canadian data handling and regional fit. SR&ED and IRAP can materially reduce early-stage risk.

If you are building vertical AI in traditional industries, I am always open to talk shop.

About the Author

Tilak Raj is the CEO & Founder of Brainfy AI, a Canadian AI company building vertical SaaS platforms across agriculture, insurance, aviation compliance, real estate, and more. He writes about practical AI implementation, vertical SaaS strategy, and building from Edmonton, Alberta, Canada.

Website: https://www.tilakraj.info Email: ceo@brainfyai.com