CV example
Data scientist CV example
Strong data science CVs connect models to money, time or risk. This example shows a data scientist who explains what each model did for the business, not just which algorithm it used.
Below you'll find a full example CV (also called a résumé), summaries for different experience levels, the skills and keywords employers search for, stronger bullet points, and the interview questions to prepare for.
Data scientist CV summary examples
Your summary is the first thing a recruiter reads. Two or three sentences: who you are, your strongest evidence, and what you want next. Here is the example CV's summary, plus versions for other levels.
Example CV
Data scientist with 4 years of experience building forecasting and recommendation models in Python. I take models from notebook to production and measure them against business outcomes, not just accuracy scores.
New graduate
M.Sc. graduate in data science with a thesis on demand forecasting and 2 Kaggle competition medals. Comfortable with Python, SQL and scikit-learn, and keen to put models into production.
Senior
Senior data scientist with 8 years leading pricing and personalisation work. Built the experimentation platform used by 6 product teams and mentor 4 data scientists.
Skills to put on a data scientist CV
Hard skills
- Python (pandas, NumPy, scikit-learn)
- SQL
- Statistics and experiment design
- Machine learning (supervised and unsupervised)
- Deep learning, if the role needs it (PyTorch or TensorFlow)
- Deploying and monitoring models (MLOps)
- Data visualisation
Soft skills
- Translating business problems into data problems
- Communicating uncertainty
- Pragmatism over perfection
- Collaboration with engineers
Show soft skills through your bullet points rather than listing them on the CV itself.
Keywords recruiters and ATS look for
Many employers screen CVs with an applicant tracking system (ATS) that matches words from the job ad. These come up often in data scientist postings. Use the ones that are true for you, in the same wording as the job you're applying for. To compare your CV with a specific job ad, use the free CV keyword checker, or have Litjob tailor your CV to the job description for you.
- data scientist
- machine learning
- Python
- SQL
- statistics
- predictive modelling
- A/B testing
- scikit-learn
- deep learning
- MLOps
- NLP
- forecasting
For live figures, see the skills most requested in Data and AI job ads, updated weekly from thousands of postings.
Bullet points: before and after
Strong bullet points say what you did, how big it was and what changed. Compare:
WeakBuilt machine learning models.
StrongerBuilt a churn model that identifies at-risk customers 30 days early, now used by the retention team.
WeakUsed Python for data analysis.
StrongerAnalysed 3 years of pricing data in Python and recommended changes that lifted margin by 2 points.
WeakWorked on NLP.
StrongerFine-tuned a text classifier that routes 70% of support tickets automatically.
WeakDeployed models.
StrongerMoved 4 models from notebooks to a scheduled pipeline (Airflow, MLflow) with weekly drift checks.
Only use numbers you can explain in an interview. An honest estimate (“about 30%”) is better than a made-up exact figure.
Tips for writing a data scientist CV
Measure models in business terms
Accuracy and AUC matter to other data scientists. Revenue, cost, time saved and risk reduced matter to the person hiring you.
Show you can ship
Many candidates only show notebooks. One bullet about deployment, monitoring or working with engineers sets you apart.
Keep projects relevant
Choose portfolio projects close to the company’s domain. A forecasting project fits retail; a fraud project fits fintech.
List methods you can defend
Interviewers will ask how a method works. Only list techniques you can explain and have used for real.
Common data scientist interview questions
Your CV gets you the interview, and interviewers will ask about what's on it. Prepare for these:
- How would you explain your model to a non-technical manager?
- Focus on what it predicts, how good it is in business terms, and its main limitation.
- How do you handle imbalanced data?
- Mention resampling, class weights, choosing the right metric (precision, recall) and threshold tuning.
- Your A/B test shows a 2% lift. Do you ship it?
- Talk about significance, sample size, practical impact, guardrail metrics and novelty effects.
- Tell me about a model that did not work.
- Show what you learned and how you found out, ideally before it hurt users.