← Atlassian Interview Insights

Atlassian·Machine Learning Engineer·Onsite - Multi Round·Senior

Senior
Apr 2026

Summary

Three-part ML interview at Atlassian covering past project background, domain knowledge in the candidate's stated areas, and a system design problem around building a cross-product RAG system. The design portion was the real meat of it and went pretty deep.

Questions Asked (4)

Q1

Walk us through your past ML projects and what your contributions were.

Technical Trade-offsAdaptability & Ambiguity
Author's notes

Pretty standard opener but it set the tone for everything after.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Select 2-3 ML projects that showcase different skills (e.g., modeling, deployment, experimentation) and align with the role's focus on technical trade-offs and adaptability. For each, briefly describe the problem, your specific contributions, the technical decisions you made, and the measurable impact. Emphasize how you navigated ambiguity and balanced trade-offs.

Pro tip: Quantify your impact with metrics (e.g., 'improved model accuracy by 15%') and explicitly state what you would do differently now, showing growth and self-awareness. This demonstrates maturity and a learning mindset.

1. Set the context

Briefly describe the project's goal, team size, and your role. Keep it concise to focus on your contributions.

2. Highlight your contributions

Detail the specific tasks you owned, such as data preprocessing, model selection, feature engineering, or deployment. Use 'I' statements to clarify your impact.

3. Discuss technical trade-offs

Explain key decisions, alternatives considered, and why you chose a particular approach. Mention constraints like latency, cost, or data quality.

4. Show adaptability to ambiguity

Describe how you handled unclear requirements, changing data, or shifting priorities. Emphasize iterative experimentation and learning.

5. Share results and lessons

Quantify the project's impact (e.g., accuracy, revenue, efficiency) and reflect on what you learned or would improve.

Key Points to Mention

  • Specific ML techniques and tools used (e.g., TensorFlow, PyTorch, scikit-learn)
  • Quantifiable outcomes (e.g., model performance metrics, business impact)
  • Trade-offs between model complexity, interpretability, and performance
  • How you handled ambiguous or evolving requirements
  • Collaboration with cross-functional teams (e.g., product, engineering)
  • Lessons learned and how you applied them to future projects

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

Deep-dive questions on the ML areas you listed as expertise on your resume.

Technical Trade-offsAlgorithms & Data Structures
Author's notes

This part is humbling if you've been loose with your resume.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

For each ML area listed on your resume, prepare a structured deep-dive that covers the problem, your approach, trade-offs, and impact. Use the STAR method to narrate specific projects, and be ready to discuss technical details, alternatives, and lessons learned. Tailor your answers to Atlassian's context by emphasizing collaboration, scalability, and data-driven decision-making.

Pro tip: Don't just list technologies; demonstrate deep understanding by discussing trade-offs and failures. Interviewers at Atlassian value humility and learning from mistakes as much as successes.

1. Clarify the Area

Briefly define the ML area and its relevance to the role, ensuring alignment with the interviewer's expectations. Ask if they want a specific project or a general overview.

2. Describe the Problem and Data

Explain the business problem, dataset characteristics, and constraints. Highlight any data challenges and how you addressed them.

3. Detail Your Approach

Walk through your methodology, including model selection, feature engineering, and evaluation metrics. Justify your choices and mention alternatives considered.

4. Discuss Trade-offs and Results

Articulate the trade-offs (e.g., accuracy vs. interpretability, latency vs. performance) and quantify the impact of your solution. Share metrics and business outcomes.

5. Reflect and Connect

Summarize key learnings, what you would do differently, and how this experience applies to Atlassian's challenges. Invite follow-up questions.

Key Points to Mention

  • Specific ML algorithms and techniques (e.g., deep learning, ensemble methods) with rationale for selection
  • Trade-offs between model complexity, interpretability, and computational efficiency
  • Data preprocessing, feature engineering, and handling of imbalanced or noisy data
  • Evaluation metrics beyond accuracy (e.g., precision/recall, AUC, business KPIs)
  • Deployment considerations: scalability, monitoring, and maintenance of ML models
  • Collaboration with cross-functional teams and communication of technical concepts to non-experts

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q3

Design a cross-product RAG system that spans two products, for example Jira and Confluence. How would you handle ingestion, change-data-capture, chunking, embeddings, retrieval, permissions, and online evaluation?

System DesignTechnical Trade-offsAPI & Integrations
Author's notes

This was the bulk of the interview and honestly the most interesting part.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying requirements and constraints, then walk through the end-to-end pipeline for each component (ingestion, CDC, chunking, embeddings, retrieval, permissions, evaluation), emphasizing how cross-product integration and permission enforcement are handled. Highlight trade-offs and justify design choices with respect to scalability, latency, and security.

Pro tip: Emphasize that permissions must be enforced at query time using the user's identity, not just at ingestion, to avoid stale access. Also, mention that chunking strategies should be product-aware (e.g., Jira issues vs. Confluence pages) to preserve context.

1. Clarify Requirements and Constraints

Ask about scale (number of documents, QPS), latency requirements, permission models, and cross-product use cases. This ensures the design meets actual needs.

2. Design Ingestion and CDC

Describe how to ingest data from both products using APIs/webhooks, and implement change-data-capture to keep the index up-to-date. Consider batch vs. streaming, and handling deletes.

3. Chunking and Embedding Strategy

Explain product-specific chunking (e.g., by issue or page section) and embedding generation. Discuss model choice, dimensionality, and storage in a vector database.

4. Retrieval and Permission Enforcement

Detail hybrid retrieval (keyword + vector) and how to enforce permissions at query time by filtering results based on the user's access rights across both products.

5. Online Evaluation and Monitoring

Propose metrics (e.g., recall, latency, user feedback) and A/B testing to continuously evaluate and improve the system.

Key Points to Mention

  • Use webhooks/APIs for real-time CDC and handle deletes/updates gracefully.
  • Product-aware chunking to preserve context (e.g., Jira issue descriptions vs. Confluence page sections).
  • Vector database with metadata filtering for permissions and product type.
  • Query-time permission checks using user identity and ACLs from both products.
  • Hybrid retrieval combining BM25 and dense vectors for better recall.
  • Online evaluation with metrics like click-through rate, user feedback, and latency.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q4

How would you approach online evaluation for a RAG system like this in production?

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

Got folded into the design question but felt like its own thread.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining clear offline and online metrics that capture retrieval quality, generation quality, and end-user satisfaction. Then describe a layered evaluation strategy: offline benchmarks for regression testing, online A/B tests for causal impact, and continuous monitoring with guardrail metrics. Emphasize the need for human-in-the-loop evaluation and feedback loops to handle the open-ended nature of RAG outputs.

Pro tip: Frame online evaluation as a product experimentation problem, not just a model metric problem—tie every metric to a business KPI (e.g., task success, time-to-resolution) and highlight how you'd handle novelty effects and long-term holdbacks.

1. Define success metrics and guardrails

Identify primary metrics (e.g., answer relevance, factuality, user engagement) and guardrail metrics (e.g., latency, cost, toxicity). Map them to business outcomes like customer satisfaction or support ticket deflection.

2. Establish offline evaluation pipeline

Create a golden dataset with diverse queries and reference answers. Use automated metrics (e.g., RAGAS, faithfulness, context precision) and periodic human review to benchmark model changes before deployment.

3. Design online A/B tests

Randomize users into control (current system) and treatment (new RAG variant). Measure primary and guardrail metrics with sufficient power, and consider interleaving or switchback designs for low-traffic scenarios.

4. Implement continuous monitoring and feedback loops

Deploy real-time dashboards for key metrics and set alerts for anomalies. Collect implicit (clicks, dwell time) and explicit (thumbs up/down) user feedback to detect drift and surface failure cases.

5. Iterate with causal inference and long-term holdouts

Use techniques like CUPED to reduce variance, and maintain a long-term holdout group to measure sustained impact. Analyze segment-level effects to ensure fairness and identify improvement areas.

Key Points to Mention

  • Offline metrics: retrieval precision/recall, answer faithfulness, context relevance (e.g., RAGAS, TruLens).
  • Online metrics: task success rate, user engagement (click-through, dwell time), deflection rate, CSAT.
  • A/B testing best practices: randomization unit, sample size calculation, novelty effects, guardrail metrics.
  • Human evaluation: periodic audits, crowd-sourced ratings, expert review for high-stakes queries.
  • Monitoring and drift detection: data drift, concept drift, feedback loops, alerting.
  • Causal inference methods: CUPED, switchback tests, long-term holdouts, segment analysis.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.