Senior Principal Machine Learning Engineer, Search Platform

Atlassian develops software products and solutions that help teams collaborate and work effectively. The Senior Principal Machine Learning Engineer will lead the technical direction and architecture for ML-based search serving, improving search quality, latency, cost, and reliability at scale. The role also involves driving search innovation, optimizing models, defining ML systems standards, mentoring engineers, and partnering across teams.

Responsibilities:

• Set technical direction for ML-based search serving: ranking models, retrieval architectures, inference pipelines, and serving infrastructure
• Drive measurable improvements across search quality, latency, serving cost, and system reliability through rigorous experimentation and principled engineering
• Identify and pursue moonshots — high-ambition, calculated bets on search techniques that deliver step-change improvements
• Lead model optimisation end-to-end: quantisation, distillation, batching strategies, hardware-aware inference, and latency/accuracy trade-offs
• Define and enforce ML systems standards: model evaluation, shadow traffic testing, rollout safety, and production observability
• Mentor and elevate senior engineers across Search Serving; raise the technical bar through design reviews, architecture decisions, and hands-on guidance
• Partner across teams - Search Quality, ML Platform, and product - to align roadmaps and unblock high-impact work
• Translate ambiguous problems into clear technical bets with measurable success criteria

Requirements:

• 12+ years of engineering experience, with significant depth in ML systems and production ML infrastructure
• Proven track record designing and shipping low-latency ML serving systems at scale (sub-100ms ranking, retrieval, or inference pipelines)
• Deep expertise in information retrieval and search: dense/sparse retrieval, neural reranking, hybrid search, learning-to-rank
• Strong command of model optimisation techniques and experience with large-scale serving infrastructure: model serving frameworks (Triton, TorchServe, vLLM or equivalent), GPU/CPU optimisation, autoscaling
• Track record of technical leadership without authority - influencing architecture and decisions across team and org boundaries
• Demonstrated ability to identify high-leverage research directions and drive them from prototype to production
• Experience with online experimentation and rigorous evaluation frameworks for ML systems
• Strong communication skills - able to distill complex trade-offs into crisp decisions for technical and non-technical audiences
• Experience with enterprise search, multi-tenant serving, compliance-constrained environments (FedRAMP, isolated cloud), or cloud-native ML on GCP/AWS