Real-time voice AI for Canada’s future. Built in Vancouver. Models, engine, platform and products — one stack, one team.
As AI agents start talking to people, voice becomes infrastructure. Infrastructure is something a country should be able to build for itself. Alebex is building it here.
*Alebex Canadian TTS, coming soon. Everything else shown is built and running today.
Speech recognition, turn intelligence, the agent runtime and validation are in-house, on our own inference and orchestration layer. Carrier connectivity stays with partners at the edge.
When more of the path is under our control, changes to timing or tone ship in days, not on a vendor’s roadmap.
Engineered toward sub-800ms responses.
An engineering target, not a measured production figure. Sub-400ms perceived with the dual-model response system, in development.
Brighton College and other Canadian colleges run qualification and booking on it today.
Serving 140+ institutions, they integrate voice agent workflows into their own products.
Their agents, their customers, our engine. API, CLI and MCP server.
140+ is the partner’s footprint, not Alebex deployments.
We began on international providers. Then we built the core ourselves, in Vancouver. Every model we ship adds capability that stays here.
Models and intellectual property created and commercialized in Vancouver.
Applied research, engineering and commercial capability, built with the Alexander Innovation Centre.
Partners and developers ship their own voice products, for Canadian and global markets.
Canadian organizations decide where their voice data lives and how systems are deployed.
Proven in education. Built to be Canada’s real-time voice AI stack.
Built for every sector: healthcare administration · financial services · retail · logistics · hospitality · home services · government · customer support · robotics.