Yiran Li
Aspiring Data Analyst | Ex Data Analyst Intern @ FlexGen | Grad Student @ UC Berkeley
University of California at Berkeley
About Me
👋 Hi there! I’m Yiran, and I’m a data geek with a knack for solving complex problems. I’m currently honing my skills as a Master’s student in Statistics at UC Berkeley, one of the world’s leading institutions for statistical science and data analysis.
🌏 Originally from Beijing, China, I ventured across the world fueled by a deep-seated passion for data. My journey has been nothing short of thrilling, transforming me from an eager student into a data analyst and problem-solver.
🔍 Why Data Science?
For me, data isn’t just a bunch of numbers; it’s a story waiting to be told. I find immense satisfaction in diving into datasets, unraveling complexities, and emerging with actionable insights. Data science is my canvas, and Python, R, and SQL are my brushes.
🛠️ Technical Proficiency
- Programming: Fluent in Python (NumPy, Pandas, Seaborn, scikit-learn, Spark), SQL, and R (ggplot2, dplyr, tidyverse)
- Software: Proficient in AWS (S3, Lambda, SageMaker), Databricks, Git, Bash, Tableau, Microsoft Excel, JavaScript, and HTML
- Courses: Accomplished in Data Modeling & Inference, Data Science Programming, Database Application, and Machine Learning
📈 Real-world Experience
One highlight is my internship as a Data Analyst at FlexGen. I tackled the challenge of optimizing Round-Trip Efficiency (RTE) for 25 battery sites using Python. Building on this, I re-engineered the RTE calculation logic via Databricks, which now delivers real-time insights at 5-minute intervals. This technological shift set the stage for me to author a 16-page RTE methodology documentation, thereby enriching our company’s knowledge base and equipping clients for data-driven decision-making. Beyond this, I collaborated with the Analyze Team to develop Python functions that dramatically cut data-cleansing time from 2 hours to just 1 minute.
🤝 Soft Skills
Effective communication and customer engagement are as important to me as technical skills. I thrive in collaborative settings and am always eager to share my analytical expertise. I believe in a growth mindset—always learning, always evolving.
🗣️ Language
Fluent in English and native in Mandarin, I can effortlessly bridge cultural and linguistic gaps, making me an asset in diverse work environments.
🌱 What’s Next?
I’m on the lookout for full-time opportunities in data science and data analytics where I can leverage my skills to make a meaningful impact. If you’re interested in solving complex problems and telling stories through data, let’s connect!
Download my resumé.
- Machine Learning
- Quantitative Data Analysis
- Data Visualization
- Statistical Inference
M.A. Statistics, 2023
University of California at Berkeley
B.S. Statistics and Analytics, 2022
University of North Carolina at Chapel Hill
Teaching Chinese to Speakers of Other Languages (major), 2019
Shanghai International Studies University
Finance (minor), 2019
Shanghai International Studies University
Skills
Experience
- Metric Design: Developed Python scripts for RTE (Round-Trip Efficiency) calculations across 25 battery sites at diverse time intervals (annual/monthly/weekly), resulting in critical insights for site performance and optimal stability periods
- Data Pipeline Automation: Automated real-time data collection and analysis process by implementing RTE calculation logic via Databricks, achieving real-time insights every 5 minutes and a 2-hour reduction in runtime
- Supervised Learning: Trained random forest, XGBoost model of 50+ features to predict the battery SoC (State of Charge), resulting in essential insights for optimal BESS (Battery Energy Storage Systems) management
- Data Visualization: Created informative dashboards using Power BI on time-series data, presenting complex data insights in a user-friendly format, resulting in improved data communication and informed decision-making by stakeholders
- Documentation: Enriched the company’s Confluence page by publishing a 16-page documentation on RTE methodology
- Data Preprocessing: Transformed raw UK Biobank data including patients’ age, gender, clinic records, lifestyle into an optimized model-ready format, slashing features from 664 to 226 and reducing model run time by 1 hour
- Feature Engineering: Enhanced model interpretability and performance by developing a composite “Lifestyle Score” by aggregating health factors (alcohol, smoking, diet, sleep, exercise) with weighted values assigned
- Exploratory Data Analysis: Initiated a thorough EDA process, leading a group of 3 to uncover significant patterns linking patients’ diabetes status with their cholesterol, lipids, and blood pressure levels
- Supervised Learning: Employed rigorous model training and selection including LASSO, Random Forest and XGBoost, increasing the accuracy for diabetes detection by 20%
- Data Visualization: Boosted the average event ROI by 11.3% through quantitative analysis on events KPIs and visualizing key trends (locational, seasonal) in Tableau
- Team Collaboration: Led a collaborative effort between 2 stakeholders (third party vendors, internal functions) in event planning, resulting in 8 successful outcomes
