Edge AI: running AI on the device, and when to do it
Edge AI means running AI on the device itself. Definition, how it works, benefits, hardware (Jetson, Coral, Edge Impulse) and the five criteria for deciding.
We regularly share practical resources for startups
Edge AI means running AI on the device itself. Definition, how it works, benefits, hardware (Jetson, Coral, Edge Impulse) and the five criteria for deciding.
Wandercraft, Uma, Enchanted Tools, Pollen: the map of French humanoid robots, how they're used in factories and services, what they cost, and where to position yourself.
HEC Paris Incubator: entry requirements, price (€300/month), the three-batch calendar, the four selection stages, and how to get in without an HEC degree.
SEMIA, 1Kubator, pH8, Nextmed, ISU, Fluxus: Strasbourg's incubators and accelerators compared, with costs, entry criteria and the right path for your stage.
World models explained: definition, how they differ from LLMs, Yann LeCun's bet with AMI Labs, the models you can actually access (V-JEPA 2, Cosmos, Genie) and what founders can build with them.
On Luma, the host sees exactly two things before approving your registration: your public profile and your answers to the registration questions. Here's what they look at, how to fill out your profile in ten minutes, and what to do when the event is already full.
Physical AI perceives, decides, and acts in the real world. This pillar walks the full loop, from sensors to field feedback, then ties each layer back to the decisions a founder has to make: wedge, data, distribution, moat.
Antler invests before the team even exists. Day zero model, thesis, check size, equity and the application process for Antler France in 2026.
Programs, selection criteria, monthly cost, timelines and what to do if you're rejected: the founder's guide to applying to Station F.