Principal Machine Learning Engineer

Role Description

As a Principal Machine Learning Engineer, you will set the technical direction for how AI is built and shipped across Jira Service Management (JSM) - one of Atlassian's largest and most varied enterprise surfaces. You'll design and continuously improve the multi-step LLM pipelines that power our AI service experiences:

• Shaping prompting strategies
• Model selection
• Output parsing
• Orchestration to raise quality and reliability across a wide range of real-world service scenarios

Your work will go deep on the problems that separate impressive demos from dependable enterprise products:

• Build retrieval and grounding systems that surface the right context at the right moment - knowledge articles, past requests, structured data, and tool outputs
• Own the quality, latency, cost, and reliability of AI workflows running at enterprise scale across a large and diverse customer base
• Design evaluation frameworks, benchmark datasets, and scoring methods that give teams a credible, stable signal on whether AI is genuinely improving or quietly regressing

You'll run disciplined experiments that validate whether changes to models, prompts, or retrieval actually move the needle:

• Build the infrastructure to run those experiments repeatedly and confidently
• Set multi-quarter ML strategy
• Make architectural decisions that span multiple teams
• Mentor and grow emerging ML engineers
• Raise the overall bar for how AI is built and shipped across JSM

This is a role for someone who wants their fingerprints on both the deepest technical decisions and the long-term trajectory of AI in a product used by teams worldwide.

Qualifications

• 10+ years of total experience, with 5+ years of related industry experience in the MLE / data science domain
• Fluency in Python
• Solid understanding of machine learning concepts and algorithms, including supervised and unsupervised learning, deep learning, and NLP
• Familiarity with popular ML libraries such as scikit-learn, Keras/TensorFlow/PyTorch, numpy, and pandas
• Strong understanding of the end-to-end machine learning project lifecycle
• Experience designing and improving LLM-based pipelines — prompting strategies, model selection, output parsing, and orchestration across multi-step workflows
• Experience building retrieval and grounding systems (e.g., RAG) that surface relevant context to improve the quality and reliability of AI outputs
• Experience designing evaluation frameworks, benchmark datasets, and scoring methods that produce credible, stable quality signals
• Experience architecting and implementing high-performance RESTful microservices (API development for ML models)
• Familiarity with MLOps and experience scaling and deploying machine learning models in production, with ownership of quality, latency, cost, and reliability
• A track record of running rigorous experiments to validate impact, and building the infrastructure to do so repeatedly
• Focus on business practicality and the 80/20 rule; a very high bar for output quality, while recognizing the business benefit of "having something now" vs. "perfection sometime in the future"
• An agile development mindset, appreciating the benefit of constant iteration and improvement

Requirements

• Deep experience with LLM-related applications and deep-learning-based models in production, at enterprise scale
• A talent for solving ambiguous and complex problems — navigating uncertainty, breaking down hard challenges into manageable components, and developing innovative solutions
• Experience setting multi-quarter ML strategy and driving architectural decisions that span multiple teams
• Experience or passion for building cutting-edge developer tools and products powered by AI/ML