Dynamic Programming

Dynamic programming has a real but modest place in the Atlassian loop, and where you meet it depends on your pipeline. New-grad and campus candidates tend to see it first on the HackerRank online assessment, where sequence-optimization and simple DP sit alongside string parsing and greedy problems. Experienced and lateral hires meet it less often, usually as an occasional problem in the live coding round rather than a fixture. Either way it stays at the medium tier. The exotic interval-DP and bitmask puzzles that headline other companies' loops are not the Atlassian bar.

So DP earns a slice of your prep, not the center of it. Trees and the coded Code Design build are where this interview actually lives, and dynamic programming is a recurring family you keep fluent rather than grind. The recurring shapes are plain. A maximum contiguous sum, a longest increasing run, a digit string counted for its decodings. A clean one-dimensional table is far likelier than anything needing a trick you would have had to see before. And every Atlassian round runs your code. You work in your own editor, over video, under Team Anywhere. So what scores is a recurrence the tests confirm, not an elegant sketch.

The Atlassian angle

Both the online assessment and the live editor execute your code, so on a DP problem the winning order is a correct table first and the optimization second. A bottom-up solution that clears every case scores. A half-remembered O(1)-space version that drops two edges does not clear the gate, however slick it reads on the screen. Build the full dp array even when you can already see the rolling-variable collapse, get it passing, then compress it. And read the constraints for the zero, the empty input, and the single-element array the hidden suite is built to catch. If a template feels shaky underneath, the free Foundations course drills each one so this page can stay on what Atlassian actually reaches for.