Location:
Sydney, hybrid
· Full time
· Reporting to our Engineering Lead Hey there, great to meet you We're looking for a mid-to-senior full stack engineer to help us build the platform that gets Australian businesses to audit-ready climate disclosure. Our engineers are at the forefront of technology, they use the latest models and direct coding agents every day, and we're rebuilding how we work around that. Highlights
- Join a small engineering team building a product that is disrupting how mid market Australian organisations are reporting on climate.
- Own an area of our Compliance Hub
- the product our clients live in during reporting season
- Agentic development is the default here, not an experiment you have to justify. We are constantly testing and learning with the latest technology
- Package includes salary + stock options
- Sydney hybrid
- home, our Sydney office, or the coffee shop in between. We require all team members to come into the office 1 day per week.
- Regular contact with the founders and a direct line to the people making the decisions
- Reporting directly to the CTO.
- Work with purpose: help Australian business actually understand and action climate risk, not just talk about it About Trace Australia's largest companies now have to report their climate risk and emissions under the Australian Sustainability Reporting Standards (ASRS), to a standard their auditors will sign off on. Most of them are not ready. Trace exists to make that possible. We combine AI-powered technology with expert consulting so businesses can produce audit-ready climate disclosures without hiring an army of consultants. We're a VC-backed Australian climate tech company, and we're currently growing faster than at any point in our history. The regulation is only getting broader from here.



How we build We're honest about this bit because it shapes the job. Our platform is a Compliance Hub plus a growing set of AI agents that run end-to-end reporting workflows
- the kind of work that used to take a consultant weeks. So the way we build mirrors the thing we're building. Agentic workflows are a normal part of daily execution here, well beyond autocomplete. Our stack is .
NET 9 on AWS Lambda for the backend, Angular (TypeScript, RxJS) on the front end, and Postgres on Aurora Serverless v2 underneath it all with API Gateway, Step Functions, SQS/SNS/EventBridge, OpenSearch and ElastiCache (Valkey) rounding out the serverless side, Cognito for auth, and Bedrock powering our in-product AI agents. Infrastructure is all Terraform, CI/CD runs through Bitbucket Pipelines, and EF Core is our ORM of choice on the .
NET side. You don't need hands-on experience with this exact stack if you've worked deeply with the equivalent (say Azure instead of AWS, or a different serverless/IaC setup), that counts just as much. Your responsibilities
- Ship features across the stack in our Compliance Hub
- from data ingestion and calculation through to the client-facing workflows that make reporting season survivable
- Take a problem from a loose product conversation to something in production, without needing the spec handed to you
- Direct coding agents across your work: scope the context, decompose the problem, review hard, and stand behind the result




- Build and improve the internal tooling, prompts, harnesses and tests that make the rest of the team faster
- Help us work out where AI agents genuinely belong in our product, and where a deterministic system is the honest answer
- Review your teammates' work — human-written and agent-written alike — and help set the standard for what "valuable" looks like here
- Work directly with our consulting team, who are the domain experts and your best source of truth on what the standards actually require What we'll love about you As an engineer
- 4+ years building and shipping production software, with a track record of things you owned end to end
- Genuinely full stack
- comfortable in the front end, the back end and the bit in the middle where the data lives
- You've maintained and debugged your own work in production, not just handed it over
- You can explain a technical trade-off to a non-technical person without making them feel stupid — you'll be doing this with our consultants and our clients
- Comfortable with ambiguity, tight timelines and a roadmap that moves. This is a startup, we move quick. With AI
- Multi-step agentic workflows are already your default way of building software, well beyond Copilot-style autocomplete. You're doing this now, not open to trying it
- Valuable judgement about delegation: what to hand to an agent, what to scope down first, and what to write yourself
- You review AI-generated code with the same discipline you'd apply to a junior teammate's PR, and you understand and own every line you submit
- You've built something to make your own workflow faster — a custom command, a skill, a CLI, an MCP server, a set of evals
- You can tell us about a time an agen

📌 Senior Software Engineer (Sydney)
🏢 trace
📍 Sydney