Deep Learning
Samsung Stock Price Forecasting
LSTM and GRU models in TensorFlow forecasting Samsung stock prices (R² = 0.95 on GRU); 10-day predictions that advised against investment ahead of an expected decline.
Product Manager · Data Scientist
Where product meets data science.
I'm Jaewon, a Product Manager at DocuSign with the toolkit of a data scientist. I pair product judgment with modeling, experimentation, and analytics to turn ambiguous problems into clear, measurable decisions.
Product Management · Roadmapping
Experimentation · A/B Testing
Python · SQL
Machine Learning · scikit-learn · XGBoost
Deep Learning · TensorFlow · PyTorch
Tableau · Power BI · Looker
Snowflake · dbt · Airflow
Causal Inference · Forecasting
About
Product thinking, powered by data science.
I studied Data Science at UC Berkeley and now work as a Product Manager at DocuSign, working across both the HR and Legal teams. My favorite work lives where product judgment meets evidence: framing the right problem, then using SQL, experimentation, and machine learning to answer it.
Before product, I trained as a data scientist and analyst across hardware, energy, and SaaS, shipping ML models, dashboards, and automation that saved real time and money. That background is how I think today: I care about products that are measurably better, not just shipped.
Whether the title says product manager or data scientist, the throughline is the same: I use data to make better decisions. A few of my projects are below.
Currently
Product Manager
DocuSign · HR / Legal
Education
B.A. Data Science
UC Berkeley · GPA 3.9
Based in
San Francisco Bay Area
Open to conversations
Focus
Product · Data Science · Experimentation
Data-informed decisions
Experience
Where I've worked.
Product ManagerDocuSign
Product Manager on the HR/Legal team, working across both HR and Legal, pairing product ownership with hands-on data science.
- Built an employee attrition model (LightGBM) that surfaces the highest-risk 15% of teams, giving HR an early signal to prioritize retention.
- Led an automation integration project connecting 10+ internal applications, cutting manual HR and Legal intake handoffs by ~40% and speeding turnaround ~30%.
- Defined requirements and success metrics for HR and Legal intake workflows with Legal, Engineering, and Data, lifting on-time SLA compliance above 95%.
- Built SQL and Tableau dashboards giving leadership real-time visibility into intake volume, SLAs, and bottlenecks.
Market Research Analyst InternAdvanced Energy
Ran competitive analysis on efficiency, pricing, and market positioning. Automated workflows with Excel VBA and Python, cutting analysis turnaround time by ~90%, and delivered insights that guided product differentiation.
Product Manager InternDocuSign
Built Tableau dashboards consolidating 50M+ records via SQL for leadership visibility; developed a Python churn/renewal model (92% accuracy) that cut manual reporting ~80%; analyzed 40K+ sales sequences to lift conversion ~15%.
Data Scientist InternMKS Instruments
Deployed ensemble ML models for defect scoring, saving ~$750K annually in scrap costs; migrated 20+ dashboards from Tableau to Power BI; automated warranty trend reporting.
Python & Mathematics TutorTublet
Taught 100+ students Python, statistics, and calculus. Earned the Honorable STEM Tutor Certificate (top 1%).
Portfolio
Projects.
Personal, end-to-end projects across exploratory analysis, forecasting, and data modeling/engineering.
Experimentation
E-Commerce Landing Page A/B Test
A/B test using a two-sample Z-test for proportions, showing the redesigned page did not significantly lift conversion versus the original.
Regression
California Housing Cost Modeling
EDA and a random forest regression to predict housing prices, with preprocessing, evaluation, and hyperparameter tuning, reaching R² = 0.80.
SQL · Analysis
COVID-19 Statistical Analysis
PostgreSQL and advanced SQL (CTEs) for multivariate analysis of COVID-19 data, identifying the most infectious countries and a −0.751 GDP-to-infection correlation.
SQL · Analytics
Data Science Job Market Analysis
A PostgreSQL pipeline surfacing the highest-paying Data Scientist roles and the skills behind them, balancing frequency and pay to score "optimal" skills to learn.
Data Cleaning
SQL Data Cleaning: Nashville Housing
A reproducible SQL cleaning pipeline that de-duplicates, standardizes fields, and enforces types, producing analysis-ready tables with clear documentation.
Web Scraping
Amazon Rating Tracker
A BeautifulSoup + Requests scraper capturing product title and average rating, writing to CSV on a set schedule (e.g. every 12 hours) to track rating trends over time.
Portfolio
Dashboards.
Interactive dashboards built to explore real datasets and deliver insights.
Tableau
London Bike Rides
Data prepped in Python pandas and presented as a multi-view Tableau dashboard.
Details
- Five visualizations including moving-average charts with parameters.
- Filters for detailed exploration.
- Heatmap of ride counts by wind speed and temperature.
- Rides by weather and rides by length.
Power BI
Data-Professionals Survey
600 responses summarized with interactive visuals in Microsoft Power BI.
Details
- Treemap filtering by country.
- Average salary by job title (stacked bars).
- Favorite languages per role (stacked columns).
- Difficulty ratings (donut), salary & work-life balance (gauges).
- Salary by age and gender (line charts).
Tableau
Seattle Airbnb
Geographical cost analysis with multiple breakdowns of Seattle Airbnb data.
Details
- Relationship between bedrooms and cost per night.
- Average cost by zip code.
- Total weekly Airbnb revenue in Seattle.
- Listings grouped by bedroom count.
Excel
Worldwide Bike Sales
Cleaned worldwide bike-sales data in Excel and presented it as an interactive dashboard.
Details
- Removed duplicates; populated marital-status column.
- Age-bin column via nested IF statements.
- Pivot tables with slicers for filtering.
- Income vs purchases, purchases by age, and commute distance vs purchases.
Contact
Let's connect.
Open to conversations about product, data, and everything in between. Reach out anytime.
Reach me
Email jaewon.shim@berkeley.edu