Canadian AI expert Raquel Urtasun launches Waabi, which aims to build a self-driving platform that's capable of complex reasoning, with $100M CAD Series A
Meagan Simpson / BetaKit :
Context & Ripple Effects
Raquel Urtasun is leaving the lab Uber built around her — she was hired in 2017 to lead its Toronto AI research operation — to found Waabi with a thesis that self-driving should be a reasoning problem solved in software rather than a sensor-stacking problem. The $100M CAD Series A is one of the larger debut rounds in Canadian AI.
Waabi enters a debate already running in the field: Wayve has raised on the argument that machine learning alone, without lidar-class hardware, can teach a vehicle to drive. Urtasun's bet on complex reasoning puts her squarely in that camp, and her academic pedigree plus Uber ties give it immediate credibility.
First-order effects
- Waabi starts life unusually capitalized for a seed-stage autonomous player, letting it pursue a simulation-and-reasoning platform instead of a hardware fleet build-out from day one.
Second-order effects
- Wayve's lidar-skeptical fundraising line gets a high-profile validator, sharpening the split between learning-first startups and the sensor-heavy incumbents competing for the same investor capital.
Third-order effects
- If the reasoning-first approach proves out at scale, the autonomous stack could consolidate around learned models rather than per-vehicle sensor suites — though whether that architecture generalizes beyond controlled domains remains the open question the funding alone does not answer.
The trend: Autonomous driving is splitting along architectural lines, with capital increasingly backing learning-based reasoning platforms over hardware-intensive sensor stacks.