Atlassian Takeover 2026
Atlassian are continuing their association with SydJS for another superb year, and you get the rewards.
One month a year they go above and beyond — clearing the runway and making room for an Atlassian Takeover. And this year their team arrives fresh from EngFest, their twice yearly internal conference.
Fresh in from New York, Justin Talbott will share how the Loom team went from a meeting recording failure rate of ~6% to under 0.1% — a story of escape rooms, browser automation, and eventually breaking free.
Shrey Somaiya will show us how static analysis can reduce your workload from weeks to seconds.
And Coby Plain will reveal "Good Developer Habits" that are bad at Atlassian scale.
It's set to be a GREAT night!
Tuesday, 3 March 2026
06:00 PM — 08:30 PM
Canva Headquarters
110 Kippax St, Surry Hills NSW 2010
Atlassian Takeover 2026
Tuesday, 3 March 2026
06:00 PM — 08:30 PM
Canva Headquarters
110 Kippax St, Surry Hills NSW 2010
Atlassian are continuing their association with SydJS for another superb year, and you get the rewards.
One month a year they go above and beyond — clearing the runway and making room for an Atlassian Takeover. And this year their team arrives fresh from EngFest, their twice yearly internal conference.
Fresh in from New York, Justin Talbott will share how the Loom team went from a meeting recording failure rate of ~6% to under 0.1% — a story of escape rooms, browser automation, and eventually breaking free.
Shrey Somaiya will show us how static analysis can reduce your workload from weeks to seconds.
And Coby Plain will reveal "Good Developer Habits" that are bad at Atlassian scale.
It's set to be a GREAT night!
Talks
From Weeks to Seconds: How Static Analysis Powers Developer Productivity at Scale
by Shrey Somaiya
This talk shares how the Jira Frontend Platform team automated what has previously taken months, if not years of manual work — into a suite of automated codemods that runs in seconds. We changed hundreds of thousands of files across hundreds of automated pull requests in multiple initiatives across an 11+ million line codebase — all without breaking a thing!
The secret? A Rust-powered static analysis engine that understands JavaScript/TypeScript dependencies at scale. By accurately mapping how files connect to each other, we harnessed a system that can safely transform code across the entire codebase automatically.
The talk covers the problem (why manual refactoring doesn't scale), the solution architecture (how Facts Map works and why we chose it), and the real-world application across multiple initiatives — past, present, future. You'll leave with practical patterns for automating large-scale code transformations, understanding when static analysis becomes a force multiplier for productivity, and concrete evidence that investing in developer tooling compounds into massive velocity gains.
"Escape Room Software Engineering" and how Loom Meeting Recordings eventually broke out to greatly improve reliability
by Justin Talbott
Throughout the last year, the Loom Meeting Recording team has had to react to sudden, unexpected changes to our supported video platforms. At the time of the feature's inception, there were no official real-time APIs to consume meeting video and audio, so we were forced to use browser automation to record meetings.
Because these integrations are unofficial, the video platforms have no obligation to alert us of new platform changes or help us when we encounter issues. When new issues arise, we were often left hunting for clues in auto-generated HTML, minified JavaScript files, and guesses as to what their server might be doing. We've internally coined the term "Escape Room Software Engineering".
The talk covers this concept, explains our methods of confronting this type of problem solving, and describes how we went from a meeting recording failure rate of ~6% to under 0.1% in the last quarter. It also emphasises the strong preference to avoid this type of software maintenance when possible, and describes our adoption of real-time APIs that Google and Zoom currently have in developer preview that will negate the need to solve these types of issues in the future.
"Good Developer Habits" that are bad at Atlassian scale
by Coby Plain
A look at a number of common conventions that developers and managers often consider good practice but can actually be detrimental to codebase and product health. We'll explore why, for larger codebases, these seemingly innocuous (or even outright good) things have unexpected consequences — and offer alternatives to achieve similar goals.