Job Description
Software Engineer | 3D Processing Platform
Sydney | Core Framework | Python | Real Engineering
This is not web.
Not CRUD.
Not “data engineer but with a 3D buzzword”.
This is a core processing platform role sitting at the heart of a production 3D pipeline.
If you like owning real systems, shaping architecture, and improving actual algorithmic throughput — keep reading.
The Role
You’ll take ownership of a Python-based processing framework that underpins large-scale 3D workflows.
This is platform engineering in the literal sense: core algorithms, performance, system design, reliability.
You will:
-
Build and evolve a Python processing framework (NumPy-heavy, algorithmic)
-
Work deeply in 3D domains (point clouds, geometry, spatial reasoning)
-
Integrate large-scale processing into cloud infrastructure (without becoming “just infra”)
-
Partner with ML, DevOps and 3D web architecture engineers
-
Coordinate execution with a high-performing offshore team
-
Improve throughput, quality, observability and stability across the pipeline
You're not building APIs for dashboards. You're shaping the engine.
What We’re Looking For
Core:
-
5+ years software engineering experience
-
Strong Python (comfortable with NumPy and scientific stacks)
-
Solid cloud fundamentals
-
Genuine 3D conceptual reasoning ability
-
Docker + Linux fluency
-
Clear communicator across time zones
Nice to Have:
-
AWS or GCP
-
CI/CD exposure
-
Prefect or Ray
-
MongoDB
-
Experience improving processing pipelines at scale
Backgrounds That Translate Well
-
3D processing / point cloud systems
-
Photogrammetry pipelines
-
Geospatial (true 3D, not just maps)
-
Computer vision / morphology
-
Robotics perception stacks
-
Scientific computing
-
Simulation engines
-
Graphics-adjacent processing frameworks
If you’ve worked on perception for drones, SLAM pipelines in robotics, reconstruction from multi-view imagery, or scientific simulation engines, this is the right flavour.
What Success Looks Like
First 90 Days
-
Understand framework architecture and core algorithms
-
Take ownership of key modules
-
Establish cadence with local + offshore teams
-
Ship meaningful fixes or performance wins
At 12 Months
-
Framework is measurably more stable and scalable
-
Throughput and quality improvements are quantifiable
-
Documentation isn’t tribal knowledge anymore
-
Cross-team trust is high
-
Offshore execution runs smoothly and predictably
Who This Suits
Someone who:
-
Gets satisfaction from making core systems faster and cleaner
-
Likes algorithmic problem solving in spatial domains
-
Wants ownership without bureaucracy
-
Is comfortable sitting between ML and infrastructure without being boxed into either
If you’re bored of shipping endpoints and want to work on the engine room of a 3D platform, this is that move.
Apply direct or reach out to Toby Tank: toby@theonset.com.au