.webp)
Resumes
August 13, 2026
Resume Example
Use these ATS-friendly resume examples and templates to showcase the models you shipped and the metrics they moved as a Data Scientist.
Helena Sorensen
(206) 555-0174 ◇ Seattle, WA
helena.sorensen@example.com ◇ linkedin.com/in/helena-sorensen ◇ helenasorensen.com
Objective
Data scientist with 6 years building models that ship, currently owning churn and recommendation models for a subscription product with 4 million users. Cut monthly churn 18% with a retention model now driving 3 automated interventions, and lifted recommendation click-through 27% across 2 model generations. Measures work by the product metric that moved, not by model accuracy alone.
Education
M.S. in Statistics, Puget Sound Crown University 2017 – 2019
Thesis on causal inference for observational product data.
B.S. in Mathematics, Cascade Line University 2013 – 2017
Coursework in Probability, Linear Algebra, Optimisation, and Computer Science.
Skills
Modelling
Churn and propensity modelling, recommendation systems, gradient boosting, causal inference, uplift modelling, feature engineering
Experimentation
A/B test design and analysis, power calculation, sequential testing, metric definition, guardrail metrics, experiment review
Production
Model deployment and monitoring, feature pipelines, drift detection, retraining schedules, working with engineering on serving
Tools & Platforms
Python, SQL, scikit-learn, XGBoost, PyTorch, Spark, Airflow, MLflow, dbt, Snowflake, AWS SageMaker
Experience
Data Scientist 04/2022 – Present
Puget Sound Crown Media Seattle, WA
- Own churn and recommendation models for a subscription product with 4 million users and 12 product teams.
- Cut monthly churn 18% with a retention model that now triggers 3 automated interventions for at-risk users.
- Lifted recommendation click-through 27% across 2 model generations validated in 9 controlled experiments.
Data Scientist 07/2019 – 03/2022
Cascade Line Commerce Seattle, WA
- Built a demand forecasting model across 40,000 products that cut stockouts 22% and overstock 15%.
- Designed and analysed about 60 experiments a year for 5 product teams, with a shared metric library.
- Reduced experiment analysis turnaround from 5 days to same day through a reusable analysis pipeline.
Data Science Intern 06/2018 – 09/2018
Cascade Line Commerce Seattle, WA
- Built a fraud scoring prototype that flagged 34% more fraudulent orders than the rules-based system.
- Delivered the prototype with monitoring notes, and it was productionised 4 months after the internship.
- Presented findings to a group of 20 engineers and analysts at the end-of-summer review.
Projects
Retention Model. Built a churn model that scores 4 million users weekly and feeds 3 automated interventions. Monthly churn fell 18% in a holdout comparison, and the uplift analysis showed which intervention worked for which segment.
Recommendation Rebuild. Replaced a popularity-based recommender with a learned ranking model across 2 generations, lifting click-through 27% across 9 experiments. The second generation added freshness signals that the first had ignored.
Demand Forecasting. Built a hierarchical forecasting model across 40,000 products that cut stockouts 22% and overstock 15%, with the forecast error reported by category so buyers could see where to trust it.
Extra-Curricular Activities
- Speak at a Seattle data science meetup about twice a year, most recently on uplift modelling.
- Mentor 3 early-career data scientists a year through a women in data programme.
- Maintain an open-source experiment analysis package with roughly 600 monthly downloads.
Leadership
- Lead modelling work across 12 product teams and run a weekly model review for 6 data scientists.
- Built the shared metric library and experiment pipeline now used by 5 product teams.
- Mentored 4 data scientists to independent model ownership across 5 years.
.webp)
%20Which%20Should%20You%20Use.webp)
