SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Buildkite is hiring a Staff ML Engineer to own machine learning end-to-end within their Test Engine product. The role centers on a high-impact problem: predicting which tests will fail for a given code change, allowing teams to skip unnecessary test runs and focus on the tests that matter.
You'll own the complete ML lifecycle—from feature engineering on code changes and test history, through model training and evaluation, to building the serving layer that keeps predictions fast enough for the critical path. You'll make key trade-offs (accuracy vs. latency, handling low-confidence predictions) and build the platform so future models reach production quickly and repeatably.
This is a genuinely difficult ML problem: sparse signal, cold-start challenges on new repos, generalization across languages and frameworks, and tight latency constraints. Buildkite has already ingested billions of test runs—the raw material exists; you'll turn it into predictions.
You'll be the most senior ML person at Buildkite (~150 people, flat structure), setting direction alongside one other recently hired ML engineer. No ML org above you; high autonomy and ownership.
Ideal candidate: You've shipped and owned ML models in production (not prototyped and handed off). You've built ML that generalized across multiple use cases, not just one clever model. You're comfortable working with noisy signal and ambiguous data. Day-to-day: Python, SQL, AWS, containerized workloads, data-at-scale tools (Spark, Flink). Experience with code analysis, CI/CD systems, or ranking problems is a bonus.
Buildkite has been fully remote since 2013, with async-first culture and genuine overlap across ANZ and US-Pacific time zones. The company serves the world's leading engineering teams—your models sit in their critical path, processing billions of test runs.
