Four decades of building — and a habit of thinking like the attacker.
Hi, I'm Bassem. I started out in technology more than four decades ago, and I've never shaken the curiosity that pulled me in — the pull to take something complex apart and make it work better.
Much of that work has been in security, where the stakes are real. Years ago I was asked to design, build, and run the security system for a museum — all the bells and whistles. What stays with me is the simulations: sitting with the team, mapping exactly how a bad actor would try to compromise it — the data and the physical exhibits alike.
Thinking like the attacker before the attacker shows up is a habit I never lost — and it's the instinct underneath everything AugmentEV does.
I bought my first EV in 2013, and I've driven most of what's come along since. Then I went inside — onto the manufacturing floor at Tesla, where "production hell" wasn't a slogan so much as a plain description of the long days and longer nights we spent making electric mobility real.
I read a car the way I read a system: long, hard, and deep — past the pitch to what's actually true.
That refusal to take the story at face value is exactly why AugmentEV ships proof, not promises. I'm building it with a growing, global team — the product I always wished existed: one that actually works, that's been thought through, and that has the decency to be honest about where it is today, and where it isn't yet.
— Bassem, Founder
AI compute you can prove — on two kinds of nodes.
We run your AI agent jobs and hand back a signed, regulator-ready record of exactly what ran — today on secure nodes, and as the network grows, on the world's idle EVs.
Secure nodes, today
Jobs run on secure cloud and datacenter GPU nodes with hardware-rooted isolation. This is the product you can use now, billed per job.
Idle EVs, next
A modern EV sits parked most of the day with capable onboard compute doing nothing. The same runtime extends onto that fleet — coming online through automotive-OEM partnerships now in active discussion.
An audit trail on everything
Every job emits a post-quantum-signed Proof-of-Task-Execution (PoTE) — tamper-evident and independently verifiable, built for audit-first and regulated work.
The greenest, largest compute network may already be parked outside.
Traditional AI compute means building more carbon-intensive data centers. Our thesis is different: the world's idle, plugged-in electric vehicles already hold one of the largest latent pools of compute on earth. Orchestrated responsibly — energy-aware, thermal-aware, never touching an owner's range — that capacity becomes verifiable AI compute that pays the people who own it, with a footprint a fraction of a purpose-built data center's. We started on secure cloud and GPU nodes to prove the runtime; the fleet is where it scales.
Straight talk: we don't ask you to trust us — every job ships with a receipt anyone can check.
Let's talk compute
A direct line for each.
Member of NVIDIA Inception.