About · practice record
One engineering practice, 2014 to now.
Product engineering in Atlassian Growth, 2014–2025 — experimentation, attribution, onboarding, analytics and cross-team reporting systems; as an individual contributor, and for four of those years as an engineering manager. Independent AI products and agent-native engineering systems since 2026. Each era left a specific mechanism behind, and the ones below are still in use.
Through-line
The job has always been the same one under different names: own a product requirement, find where the straightforward implementation stops holding, follow the problem into whatever layer it actually lives in, and build the part that was missing. At Atlassian that surfaced through Growth — funnel audits, attribution formulas, experiment windows, event pipelines, onboarding and cross-team reporting. Independently it surfaces through AI product behaviour — schemas, domain validators, model contracts and workflow contracts. The last step is always the same: write down what the system is then allowed to claim.
The layers changed. The way in did not.
Practice arc · what each era left behind
2026 —
Independent AI product engineer. Codenames AI in production, combining live-product telemetry with controlled model experiments; agent-native engineering systems with explicit workflow contracts; a weekly field report on what held.
In force now — schema-first structured outputs and board-aware domain validators; model migrations run as controlled experiments rather than swapped in place.
2024 – 2025
Senior Software Engineer, Growth · Atlassian. Cross-functional growth engineering to onboard Atlassian's largest acquisition onto experimentation and growth infrastructure — acquisition onboarding, analytics pipelines, ML-backed surfaces.
Carried forward — the event-pipeline fix that restored attribution for experiments already running, instead of restarting them. Repairing telemetry without discarding the run is now the default posture.
2020 – 2024
Engineering Manager, Growth · Atlassian. Managed teams of 8–10 engineers through significant organizational and product change, and established the experiment-operations practices used by Growth teams running cross-product experimentation. Concurrently, from 2022, the engineering platform behind Atlassians in Mentoring.
Carried forward — two mechanisms. The Cross Flow attribution formula, later adopted for Growth Experiment Impact Estimation. And the experiment-operations practice: written rules for what a team may launch, restart or report — the same artifact is now a workflow contract for agents.
2019 – 2020
Senior Software Engineer, Growth · Atlassian. Experimentation initiatives, and Informed Pull Requests — surfacing security vulnerabilities in pull requests and running accessibility and performance audits on SPAs.
Carried forward — checks that run at the point of change, which is what the current hook stack does.
2014 – 2019
Graduate then Software Developer · Atlassian. Frontend and full-stack work across growth, billing, purchasing, onboarding and experimentation systems, including cross-product initiatives involving Jira, Bitbucket and Confluence. UNSW Co-op Program Scholar, Software Engineering, 2009–2013. The origin of the practice; the named mechanisms start above.
From the Atlassian Growth years
>10%
prior-approach over-attribution exposed
Cross Flow funnel audit · formula subsequently adopted for Growth Experiment Impact Estimation
Experiment operations & program scale · 2020 – 2025
Management was this practice at team scale.
Four of these years were spent managing, and the work was recognisably the same: the operating practices for teams running cross-product experimentation, delivery held steady through organizational and product change, and — concurrently, 2022 to 2025 — the engineering platform behind Atlassians in Mentoring, which was built, not sponsored.
The period ends in 2024 with a return to a Senior Software Engineer title, which is a change of title more than a change of method.
One period inside a continuous practice, not a second career track.
8–10
engineers directly managed
Atlassian Growth organization · through significant organizational and product change
3,552
matched mentors and mentees
Atlassians in Mentoring · approximately 20% of Atlassians
Current practice · 2026 —
Codenames AI is a production AI product where a reply that parses is not yet a legal move: schema-first structured outputs and board-aware domain validators decide legality, and model migrations run as controlled experiments with evaluation built into product behavior. Around it sit agent-native engineering systems — reusable workflows with explicit contracts, authority constraints and verification — and a weekly DEV field-report series.
Same mechanism, current form
- Attribution formulas
- → domain validators
- Experiment windows
- → controlled model experiments
- Experiment-operations practice
- → written workflow contracts
175+
monthly active players
Live-product telemetry · resume phrasing uses 175+ as durable floor, not the rolling monthly-active-players snapshot
Where each thing lives
- This page
- Continuity — how the career developed into the current practice. Four qualified figures, one per era; enough to show it is one practice.
- Home
- Positioning — the thesis, the two-channel method read across both systems, and the routes into every deeper surface.
- Projects
- Evidence — attribution, window variance, pipeline recovery, OKR attainment, Admin Hub activation, each with its scope.
- Articles
- Reasoning — weekly field reports on the calls made while building these systems.
- Résumé
- The full role-by-role record, on request.
Next step
Open to senior engineering roles.
Email is the fastest route; the record above is the reference.