Job Description

At 3:07 AM, a drone breaches restricted airspace. There’s no radar ping. No radio chatter. No warning. Just a low hum that vanishes into the night—until it’s too late.

Now imagine building the tech that detects it anyway.

Across battlefields, airports, critical infrastructure and conflict zones, asymmetric threats are becoming smarter, smaller, and harder to spot. What used to be science fiction is now a daily reality.
And somewhere in Sydney, a small team of engineers is quietly building the system that stops them.

The Mission

You’ll be part of the crew building the algorithmic brain behind a cutting-edge sensor fusion platform—a system that takes input from radar, RF, acoustic and optical sensors and makes real-time decisions in chaotic environments.

Your job? Design algorithms that don’t blink. That find signal in the noise. That can track, associate, and fuse sensor inputs on low-SWaP hardware in dynamic, hostile environments.

This isn’t post-processing in a lab. It’s real-world decision-making, deployed on systems that are out there right now.

The Challenge You’ll Own

Build performant sensor fusion and tracking algorithms that run on the edge

Optimise models using profiling, CUDA, multi-threading and whatever it takes

Handle uncertainty, association, clutter, and time desync—then make sense of it

Prototype fast, deploy cleanly, and automate everything you can

Be the technical benchmark for signal processing, algorithmic rigour, and edge deployment practices

The Kind of Engineer Who Thrives Here

You probably started as a pure algorithm dev—or signal processing nut—but found your way into defence, robotics, or embedded AI because you wanted to see your models make decisions. You like the noise. The edge. The pressure to get it right when there’s no second chance.

You bring:

• Deep knowledge of C++ (Python or Golang on the side is fine)
• Hands-on experience with real-time sensor fusion, probabilistic models, and tracking filters (JPDAF, MHT, Kalman stuff)
• Confidence running algorithms on constrained devices: GPU acceleration, TensorRT, Docker, Linux
• Curiosity, ownership, and a distaste for abstract research with no deployment

Bonus if you’ve worked in RF-heavy environments, deployed field-tested models, or helped scale early-stage AI/ML platforms.

Why It Matters

The team you’ll join is part of a fast-growing scale-up that started with two engineers and a crazy ambition: to build the world’s leading counter-drone tech without relying on traditional radar or jamming.

Today, that ambition is real—used in over 70 countries, across defence, government, and critical infrastructure sectors. The algorithmic foundation you lay will be directly responsible for protecting lives and sovereignty in the face of emerging threats.

If you’re tired of roles where your work never leaves the simulation—and ready to build tech that quietly outsmarts the threat before it’s even noticed—you know where to find us.

Reach out to Thaís on thais@theonset.com.au