Answer values interview behavioral questions
Company: Atlassian
Role: Machine Learning Engineer
Category: Behavioral & Leadership
Difficulty: medium
Interview Round: Onsite
You are in a **management/values** interview. Prepare behavioral answers aligned to the company’s core values.
Typical questions in this round include:
- Describe a time you **pushed back** on a decision or requirement.
- Describe a time you drove **change / improvement**.
- Describe a time you **helped other people improve** (mentoring/coaching).
- Describe a time you worked through **ambiguity**.
- Describe a **mistake** you made and how you corrected course.
Interviewers may **probe with follow-ups** if they haven’t yet heard the signal they need (scope, ownership, impact, conflict handling, learning).
Overview: This question evaluates leadership and behavioral competencies such as alignment with organizational values, ownership, communication and conflict resolution, mentorship and coaching, handling ambiguity, and learning from mistakes.
Read the full Atlassian Machine Learning Engineer interview experience this question came from
Solution
## 1) What the interviewer is really evaluating
Values/behavioral rounds typically measure:
- **Judgment:** do you make good tradeoffs and escalate appropriately?
- **Ownership:** do you take responsibility beyond your job description?
- **Collaboration & influence:** can you align others without authority?
- **Customer orientation:** do you protect user/customer outcomes?
- **Growth mindset:** do you learn from mistakes and improve the system?
## 2) Use a tight STAR structure (with an “L”)
Use **STAR + Learning**:
- **S (Situation):** 1–2 sentences, who/when/why it mattered.
- **T (Task):** what you were responsible for (be explicit).
- **A (Action):** 3–6 bullets of what you did; emphasize decisions and communication.
- **R (Result):** quantify impact (%, time saved, incidents reduced, revenue, adoption).
- **L (Learning):** what you’d repeat/avoid; what process changed.
A common failure mode is spending 70% on Situation and 10% on Actions/Results.
## 3) Build a “story bank” mapped to values
Prepare **5–6 stories** that can be reused across prompts. For each story, write:
- Your role + scope (team size, systems owned, timeline)
- Conflict/stakes (why it was hard)
- Metrics before/after
- What you learned
Then map each story to one (or more) values such as:
- Customer focus
- Teamwork
- Transparency
- Bias for action / ownership
- Continuous improvement
(Use the company’s exact wording if you know it; the mapping is what matters.)
## 4) How to answer each common prompt
### A) “Tell me about a time you pushed back.”
**Signals to hit**
- You pushed back for a principled reason (customer risk, reliability, ethics, cost).
- You proposed alternatives, not just “no.”
- You handled disagreement respectfully and aligned on a decision.
**Good action pattern**
1. Clarify goals/constraints and restate them.
2. Present data (incidents, latency, error budget, user impact, opportunity cost).
3. Offer options with tradeoffs (e.g., phased rollout, feature flag, smaller scope).
4. Align stakeholders; document decision.
**Pitfall:** sounding stubborn or political; avoid “I won,” emphasize outcome.
### B) “Tell me about a change/improvement you drove.”
**Signals to hit**
- You identified a problem others accepted as normal.
- You influenced adoption (not just built a tool no one used).
- You measured impact.
**Examples of measurable improvements**
- Reduced deployment time from 45 min → 10 min.
- Cut on-call pages by 30% via better alerts/runbooks.
- Improved conversion by X% via experimentation.
### C) “Tell me about helping others improve.”
**Signals to hit**
- Coaching approach: specific feedback + support + follow-up.
- You tailored to the person’s goals.
- You improved team throughput, quality, or morale.
**Concrete actions**
- Pairing sessions, code review coaching, writing a rubric, facilitating postmortems.
- Helping someone lead a project or present to stakeholders.
**Pitfall:** taking credit for someone else; center their growth.
### D) “Tell me about ambiguity.”
**Signals to hit**
- You created clarity: defined success metrics, risks, and a plan.
- You reduced uncertainty via small experiments/spikes.
- You communicated frequently and adjusted.
**Strong pattern**
- “I wrote a one-pager: problem statement, non-goals, assumptions, open questions, proposed milestones, and an RFC review.”
### E) “Tell me about a mistake.”
**Signals to hit**
- You own it without blaming others.
- You mitigated quickly.
- You prevented recurrence with a systemic fix.
**Great mistake stories include**
- A production incident caused by an unchecked migration.
- Mis-scoping a project and missing a key stakeholder.
- Shipping without a rollback plan.
**Must include**
- What you changed afterward: checklists, tests, alerts, canary releases, better stakeholder reviews.
## 5) Handling follow-up questions well
When probed, answer in the dimension they’re missing:
- **Scope:** “I owned X services; traffic was Y rps; 3 teams depended on it.”
- **Conflict:** “PM wanted speed; SRE wanted safety; I proposed staged rollout with guardrails.”
- **Impact:** “Result was 22% fewer support tickets and 15% faster time-to-resolution.”
- **Reflection:** “Next time I would involve Legal earlier and time-box the spike.”
If you don’t know, say what you’d do: “I don’t recall the exact number, but we tracked it via …”
## 6) Quick checklist before the interview
- Each story has **numbers** (even approximate) and a clear **before/after**.
- You can explain your decision-making and tradeoffs.
- You have 1 story demonstrating **principled pushback**.
- You have 1 story demonstrating **learning from failure**.
- You can deliver each story in **2 minutes**, with optional depth if asked.