Telling Your Story¶
"Tell me about yourself" is the most underestimated question in any interview. Candidates treat it as a warm-up. Hiring managers use it to decide in the first 90 seconds whether they want to keep listening. Your career narrative is not a summary of your resume — it is the argument for why you, specifically, are the right fit for this role.
What Interviewers Are Actually Evaluating¶
When you answer "tell me about yourself," the interviewer is assessing four things simultaneously:
- Communication clarity: can you distill a complex story into something crisp and coherent?
- Self-awareness: do you know why you made the choices you made, or are you reciting a timeline?
- Relevance: do you understand what this role requires, and can you connect your background to it?
- Energy and conviction: does your story suggest you are genuinely excited about this direction, or are you applying everywhere and hoping something sticks?
The biggest mistake candidates make is treating "tell me about yourself" as a chronological autobiography. The interviewer does not need to hear every role from 2014 to now in sequence. They need to understand: who are you, what do you do well, and why does this opportunity make sense for you?
The 90-Second Career Narrative Structure¶
A strong "tell me about yourself" answer follows a three-part arc:
Part 1 — Your professional identity (15–20 seconds) Lead with what you are, not where you came from. State your area of expertise and what you bring. This orients the listener.
"I'm a data scientist with four years of experience building predictive models in the e-commerce and fintech space. My work has mostly been on the boundary between modelling and product — I'm as comfortable talking to engineers about feature pipelines as I am talking to stakeholders about business metrics."
Part 2 — The narrative thread (45–60 seconds) Choose two or three moments from your career that illustrate a coherent progression. Not every job — the thread that connects them. What have you been building toward?
"I started in analytics at a retail company, which gave me a foundation in SQL and business reasoning — I learned to ask 'so what' before writing any query. I moved to a startup where I built my first production ML pipeline end-to-end: data ingestion, feature engineering, model training, and monitoring. That experience showed me how different building a model for a Jupyter notebook is from building one that people depend on in production."
Part 3 — Why this role, right now (20–30 seconds) Connect your narrative to the specific role. This requires research — generic closes ("I'm excited about the opportunity to grow") signal that you didn't do your homework.
"What draws me to this role at [Company] is the combination of the scale you operate at and the emphasis on causal inference — I've been actively building that skill set and I want to work somewhere it's central to how decisions get made, not a nice-to-have."
How to Frame a Career Transition Into Data Science¶
Career transitions are common in data science — many practitioners come from academia, software engineering, finance, or domain-specific fields. The key is to frame the transition as a deliberate evolution, not a pivot away from failure.
What to avoid: - "I decided I wanted to try something new" — no narrative, no direction - "Data science is the future" — everyone knows this, it says nothing about you - Extensive apology for not having a traditional data science background
What works:
Frame the thread, not the gap. Your previous domain gives you something most data scientists don't have — domain expertise. A biologist moving into biotech ML knows what the models mean. A finance professional moving into fintech ML understands the regulatory and risk context. Lead with the asset, not the absence.
Demonstrate the deliberateness of the move. Show that you have been building skills systematically: specific courses, projects, contributions, publications, internships. "I spent six months building X in my own time" is more compelling than "I took a Coursera course."
Connect the dots from where you were to where you are going. The story should feel inevitable in hindsight. "My work in operations led me to see that the biggest inefficiencies were in forecasting, which led me to study time series modelling, which led me to realise I wanted to work directly on these problems at scale."
Worked Example: Transitioning from Software Engineering to Data Science¶
Context: Four years as a backend software engineer; now applying for a data scientist role at a consumer tech company.
"I'm a software engineer who has spent the last two years making a deliberate shift into data science. My engineering background — four years building data pipelines and APIs at a fintech startup — gave me a strong foundation in what happens to models after they're built: latency constraints, infrastructure costs, data quality in production. What I was missing was the modelling side, so I started applying ML to problems in my current role: first building a simple rule-based fraud detector, then replacing it with a gradient boosting model that reduced false positives by 30%.
I've been studying statistics and ML systematically — I worked through ESL, built three end-to-end projects that are live on GitHub, and contributed to a scikit-learn documentation PR. I'm now at a point where I want to be in a role where modelling is the primary job, not a side project.
What draws me specifically to [Company] is that you're dealing with the kind of scale where the model architecture decisions matter — I've read your engineering blog and the work your team is doing on real-time feature serving is exactly the direction I want to grow in."
Why this works: - Leads with a professional identity, not a timeline - The engineering background is framed as an asset, not something to apologise for - Specific evidence of the transition (projects, concrete results, contributions) - The close is specific to the company — it requires actual research
Worked Example: A Recent Graduate with Project Experience¶
Context: BSc in Statistics; completed three independent projects and a summer internship at a small analytics firm.
"I'm a statistics graduate who has been building practical ML skills throughout university, mostly by taking on projects beyond the curriculum. My strongest area is supervised learning with structured data — I've built a customer churn prediction model using Telco data that achieved an AUC of 0.87, and a property valuation model for a student competition that placed second nationally.
My internship at [Analytics Firm] taught me how different client-facing data work is from academic work — clients don't care about your methodology, they care about the decision your analysis supports. Learning to communicate that clearly was the most valuable thing I took from that experience.
I'm looking for a role where I can work on real user data at scale and learn from experienced practitioners. The thing that stood out about [Company] to me is that your team publishes extensively — I've followed the work your research team did on counterfactual evaluation for recommendations, and that's the level of rigour I want to be working at."
Why this works: - Specific, concrete evidence (AUC 0.87, second-place competition) rather than vague claims - Self-aware about what was learned from experience, not just what was done - The close shows genuine research — not every company publishes, and referencing a specific paper signals real interest
Common Mistakes and How to Avoid Them¶
Mistake 1: Too much chronology "I went to university where I studied mathematics, then I joined Company A where I did analytics, then after two years I moved to Company B where I did more modelling..."
This is a resume reading. The interviewer has your resume. They want synthesis, not recitation. Fix: start with your professional identity today, then use your history selectively to support it.
Mistake 2: Too technical, too early Opening with the names of algorithms, frameworks, or tools before establishing context. The interviewer does not yet know what problem you were solving or why it mattered.
Fix: establish the business context (what were you trying to accomplish?) before describing the technical approach (how did you accomplish it?).
Mistake 3: Too vague "I enjoy using data to solve problems and drive business value." This is true of everyone applying for data science roles. It is meaningless differentiation.
Fix: replace every vague claim with a specific example. "I enjoy using data to solve problems" → "I built a demand forecasting model that reduced warehouse overstock by 18%, which was the first time the operations team trusted a model enough to act on it without manual review."
Mistake 4: Not knowing your own story When asked "why did you leave Company X?" or "why did you choose that project?", candidates sometimes give answers that contradict their opening narrative. Prepare your story in advance and make sure the pieces are consistent.
Mistake 5: Not tailoring to the company The close of your narrative should be specific. Generic closes ("I want to grow at a great company") tell the interviewer you have not thought about why their company, specifically, is the right place for your next step.
How to Research a Company and Tailor Your Narrative¶
Before every interview, do this in 30–45 minutes:
- Read the job description carefully: what are the three things this role most needs? Build your narrative around demonstrating those three things.
- Read recent company blog posts and engineering blog: what problems are they working on? Reference something specific in your close.
- Look at the interviewer's LinkedIn or publications: do they have a specific area of expertise that you can connect to? Mentioning it shows you prepared — not in a sycophantic way, but in a "I read your paper on X and that's related to how I've been thinking about Y" way.
- Know the company's product: use it before the interview if you can. Specific observations about the user experience are more credible than abstract praise.
- Know their recent news: fundraising, acquisitions, product launches, leadership changes. This context shapes what they need right now.
Questions to Reflect On Before the Interview¶
Work through these in writing, not just in your head. Writing forces clarity.
- If I had to describe my professional identity in one sentence, what would it be?
- What is the one thing I've built that I'm most proud of, and why does it matter beyond the technical achievement?
- What did I learn from my biggest professional failure or setback? Can I talk about this honestly?
- Why do I want to leave my current role (or, if a fresh graduate, why this specific company over others)?
- Why do I want to work at this specific company, on this specific team? What is the answer that requires actual research?
- What do I want to be doing in two years? Does this role connect logically to that future?
- What questions do I still have about this role after reading everything available publicly? (Thoughtful questions at the end of an interview are part of your story.)
Questions to Ask at the End of the Interview¶
The questions you ask reveal as much as the answers you give. Avoid questions about salary, vacation, or benefits in an early-round interview. Ask questions that signal you have thought deeply about the role.
Strong questions to ask: - "What does success look like in this role at 6 months? At one year?" - "What is the biggest unsolved problem on the team right now?" - "How does the data science team interact with product and engineering? Who typically initiates a modelling project?" - "Can you tell me about a time a model the team built didn't work in production? What did you learn from it?" - "What skills does the person who grows fastest in this role typically have?"
Questions to avoid: - "What does your company do?" (You should already know this) - "What is the data science team's tech stack?" (Fine question, but leads with tools rather than impact) - "Can I work remotely?" (Ask HR, not your interviewer)