Build AI that has to work on a Tuesday night shift.
We hire people who want their models, firmware, and pipelines running in real operations — with all the messy, satisfying constraints that come with it.
What it's like here
Ship to the physical world
Your work ends up on a plant floor or a tower site, not in a demo folder. That standard shapes everything about how we engineer.
Small teams, whole problems
You own a deployment end to end — model, hardware constraints, customer conversation — instead of a slice of a backlog.
Three offices, one team
Dover, London, and Dubai, with remote collaboration as the default and travel to deployments when the work needs eyes on site.
Learn on real constraints
Edge compute budgets, OT networks, ATEX zones, twenty-year-old machines — constraints that make engineers better, fast.
Backgrounds that thrive at Invexal
- ML / computer-vision engineers who care whether inference fits on a Jetson
- Full-stack engineers comfortable owning features from schema to pixel
- Firmware and embedded engineers who've debugged in the field
- Delivery and project leads who can walk a plant and talk to a board
No public listings right now — write anyway
Roles open as deployments land, and strong introductions get read first. Send a short note about what you've shipped to marcom@invexal.com.
Email us your work