Data Scientist Resume Examples

Junior Data Analyst — Tempus, a precision medicine technology company

May 2025 – Present · 1 yr 6 mos

• Built and maintained 6 recurring KPI dashboards in Tableau for patient intake, turnaround time, and support operations, writing the SQL behind each and documenting metric definitions with stakeholder input. Cut ad hoc reporting requests from 14 per month to 5.
• Created a Python validation workflow in pandas and NumPy (Jupyter Notebook) for weekly data pulls across CRM, billing, and clinical datasets, flagging nulls, duplicate records, and join mismatches before report refreshes. Reduced data QA time from 4 hours to 90 minutes per week.
• Assisted with measurement for 3 A/B tests on patient onboarding messaging: defined success metrics with a senior analyst, checked event-level data quality, and summarized confidence intervals and lift in SQL and Python, delivering result summaries within 2 business days of each test's completion.
• Use Git for version control across SQL, Python, and dashboard work to keep analytical workflows reproducible.

Data Scientist — Sally Beauty Holdings, a specialty beauty retailer and distributor

Mar 2023 – Present · 3 yrs 8 mos

• Built SKU-store demand forecasting models for seasonal beauty categories, reducing weekly forecast MAE by 11% versus the prior rules-based baseline by training and validating Python models on historical sales, promo, and calendar features sourced with SQL from Snowflake.
• Developed a customer segmentation model for e-commerce and loyalty audiences using scikit-learn clustering on purchase frequency, category mix, and recency features; the first 2 segmented campaigns increased email click-through rate by 1.8 percentage points.
• Designed and analyzed 5 promotion tests for digital offers and threshold discounts, defining test cells, validating event capture, and measuring lift with SQL-based experiment readouts and confidence intervals; identified 2 offers with positive incremental margin and stopped 1 underperforming test 10 days early.
• Created Tableau model monitoring dashboards tracking forecast accuracy (RMSE, MAE), campaign response, and feature refresh status, cutting monthly model review preparation from 6 hours to 2.
• Built a promotion-effectiveness dataset used across merchandising and finance, standardizing SQL transformations and partnering with analytics engineering to productionize curated features in Snowflake, reducing analyst rework by 9 hours per week.

Decision Scientist — Favor Delivery, a Texas-based delivery and e-commerce platform

Nov 2021 – Feb 2023 · 1 yr 3 mos

• Built a churn classification model for first-order customers using scikit-learn on order cadence, delivery experience, and promotion history features, reaching 0.76 AUC in validation and lifting win-back targeting precision from 0.21 to 0.33.
• Analyzed 7 A/B tests on delivery fee messaging and cart incentives, defining success metrics, checking assignment balance, and summarizing lift and confidence intervals in SQL and Python; recommended changes that raised completed checkout rate by 2.4 percentage points in the winning variant.
• Built Tableau dashboards for acquisition, retention, and delivery-time KPIs, automating recurring queries and refreshes and cutting weekly stakeholder reporting prep by 5 hours.

Senior Data Scientist — Marqeta, a card issuing and payments platform

May 2022 – Present · 4 yrs 6 mos

• Led development of a card-transaction fraud detection model for higher-risk merchant segments, raising fraud capture rate by 9 percentage points at the same manual-review volume by engineering behavioral features in SQL and Python, training XGBoost models, and calibrating thresholds with risk operations.
• Built a portfolio early-warning model for SMB program performance in scikit-learn, combining payment, cohort, and exposure features to improve 60-day delinquency recall from 0.58 to 0.71, validated by segment.
• Designed causal analyses for onboarding and pricing policy changes using A/B testing and difference-in-differences with finance and risk partners, informing 3 policy decisions including a rollout that cut first-30-day loss rate by 1.2 percentage points.
• Established post-deployment monitoring for 6 production models with MLflow and Tableau, tracking ROC-AUC, precision, KS, calibration, and stability metrics, and cut time to detect material drift from 21 days to 5.
• Mentored 4 data scientists on production modeling and review standards through reusable review checklists and documentation templates, reducing median code review turnaround from 4 business days to 2.

Data Scientist — Upgrade, a consumer lending and banking technology company

Feb 2019 – Apr 2022 · 3 yrs 2 mos

• Built an underwriting model for near-prime personal loan applicants, training gradient boosting and logistic regression models on bureau, application, and transaction features in Python and SQL, and improved KS by 7 points over the incumbent scorecard.
• Evaluated pricing and onboarding experiments across acquisition funnels, designing readouts, checking assignment quality, and summarizing lift and confidence intervals for product and finance stakeholders; one pricing variant lifted funded-loan conversion by 1.6 percentage points.
• Created a customer lifetime value segmentation for repeat borrowing and cross-sell planning, building segment rules from repayment behavior, product tenure, and engagement patterns; targeted offers built on the segments saw response rates increase by 14%.

Risk Analytics Manager — Discover Financial Services, a US card issuer and digital banking company

Jan 2016 – Jan 2019 · 3 yrs

• Rebuilt monthly portfolio monitoring for delinquency and approval trends, moving SAS-based reporting logic into SQL and Python with standardized metric checks, and cut the reporting cycle from 3 days to 1.
• Analyzed a credit line assignment policy update for new card accounts by segmenting applicant cohorts and comparing outcomes across policy bands, supporting a rollout associated with a 0.8 percentage point reduction in early charge-off rate.