What a 30-day pilot actually proves

We have built the vehicle pipeline — read the plate, check it against VAHAN, write the result to an access-controlled, exportable audit log — and verified it end to end, with the first real gate ahead of us. So we are not going to show you a wall of glossy deployment numbers. We would rather prove it on your gate.

Start in the dark

A pilot opens with roughly 48 hours of silent shadow mode. The box watches the gate and writes what it sees, but changes nothing and alerts no one. That gives both sides a baseline: how the existing setup actually performs, and how Atlas reads the same traffic, before anyone is asked to act on it.

An honest target, measured on your cameras

We commit to a first-pilot read target of at least 90% — not a marketing 99%, and not measured on a clean test rig. It is measured on your cameras, your lighting, your weather. A gate that can't read a plate through July rain at midnight isn't software for India, so we tune for the hard conditions first and report the real number.

The evidence is yours

Through the pilot, every event lands in the access-controlled, exportable audit log, and a data processing agreement is attached from day one. You can export the read log, the false-positive cases, and the confusion data per gate. The point is that you leave the pilot holding the proof, not a testimonial we wrote.

If it doesn't beat what you run today, we uninstall

That is the whole offer. We arrive within about a week with the edge box, plug into the cameras already on the wall, and run thirty days. If Atlas doesn't beat your current setup, we take the box back. No charge, no lock-in, no half-built integration left behind.

A pilot is a smaller promise than a case study, and a more honest one. It is the right size of claim for where Atlas is — and the fastest way to find out whether the gate is better with it than without it.