Robotics Software Engineer (ML & Perception)

Overview

We are an early-stage startup developing ‘bolt-on’ autonomy for the agricultural sector - hardware and software that turn existing tractors, sprayers, and harvesters into autonomous machines. We fuse advanced vision / AI navigation with rugged, field-tested hardware to deliver cost-effective and intuitive autonomy that farmers can trust.

We raised our initial funding three years ago and built our core technology with a lean team of a few engineers. We’ve proven the core technology in the field, validated market demand, and are now entering commercial trials in NZ and California - so we’re growing our engineering team over the coming months.

We’re hiring roles across software / ML, mechanical, and electronics, with multiple roles open that can be shaped around the strengths of the right candidates.

We are a small team with a lot to do, so you’ll need to be highly self-driven, able to figure things out without a detailed brief, and know when to push hard and when to be pragmatic.

Key Responsibilities

• Develop and optimise vision-based ML models (object detection, semantic segmentation, depth estimation), maintaining the datasets, training pipelines, and evaluation tools that support them.
• Explore and integrate state-of-the-art vision and ML techniques (e.g., vision transformers, self-supervised learning, temporal models) to improve robustness and generalisation under challenging orchard conditions.
• Build and maintain ML training and evaluation pipelines for perception and mapping tasks.
• Create aerial image-based classification pipelines for orchard mapping and GIS layers.
• Implement temporal scene reconstruction, fusing multi-frame data into consistent 3D world models.
• Design and optimise SLAM and visual odometry algorithms for orchard navigation.
• Classify and segment 3D point clouds derived from stereo cameras or aerial data for terrain, row, and obstacle understanding.
• Implement self-monitoring mechanisms within perception modules to ensure safety and reliability.
• Develop radar processing pipelines (e.g. for cross-traffic detection), combining outputs with vision-based perception.

Who You Are

Ideally, you’re a standout engineer - someone with strong ML expertise, fast learning ability, and the drive to solve challenging perception problems. We’re open to a range of experience levels, but ML experience is essential.

Preferred Skills / Experience

Practical things

Job details

#J-18808-Ljbffr