Frances Zhang Email & Phone Number

Scottsdale, Az, Us

Seattle, Washington, Us

Seattle, Washington, Us

Led the research and development of end-to-end AI-driven products in the merchandise space, enhancing revenue and margin efficiency, optimizing inventory health and streamlining operational processes.

Seattle, Washington, Us

Laval, Québec, Ca

Summary:Responsible for delivering advanced analytics projects that drive business results with multiple departments, including Global Marketing, Category Management, Shared Services & Finance, Global Technology, and Business Unit Leadership; Develop advanced analytics and predictive models from design through implementation in the areas of pricing and promotion, marketing, merchandising, and other areas of the business as needed;

Laval, Québec, Ca

Data Engineering:• Designed, developed and maintained automatic ETL pipeline across billion rows of transaction data using Databrick Pyspark to weekly measure test vs control performance of Localized price optimization on over 20K merchandise items in 11+ business units with lift metrics; Generated backend measurement data table to power dashboard reporting to support leadership's decision-making; Performed weekly outlier detection and Ad-hoc analytics to identify stores with abnormal sales for further investigation with BU managers.• Built data pipeline using Power App and Power Automate to enable input data to be auto-processed from data request submission, self-validation to table upload, reducing 85% manual work and processing time.Machine Learning:• Took an initiative to lead a team (3) to implement merchandise price optimization on Florida Business units over 400+ stores and 2500+ items from data preparation, elasticity modeling, price optimization to final optimized price recommendations to Florida stakeholders, generating 1.2M+ margin uplift for a year.• Worked with 2 team members to R&D and developed an enhanced elasticity modeling methodology to mitigate large price variations within store clusters while extending 50%+ more stores into Localized pricing training process; designed a pilot pipeline to test and measure performance in real practice using 3-month post observation data (performance is expected after 3 months)Mentoring:• Mentoring onboarding data scientists to quickly pick up tools, platforms and data science projects by holding multiple knowledge transfer sessions and regular check-ins.

Basking Ridge, Nj, Us

Focused on enhancing device performance, network health, user experience, and service quality through advanced data analytics, statistical testing, and ML modeling for Verizon devices, networks, and applications.Churn Prediction: Designed an automated daily workflow in PySpark/Scala to forecast monthly churn scores for 75M users, successfully identifying 40% of potential churners one month in advance.Video Streaming Assessment: Devised a scoring algorithm using Scala to evaluate both user and network video streaming experiences.Network Quality Monitoring: Crafted an end-to-end predictive system in PySpark to anticipate troubleshooting tickets for over 130M users, pinpointing 70% of users with potential network issues.ETL and Backend Maintenance: Consistently oversaw the automation of ETL processes and backend algorithms, ensuring high data integrity and reliability.

Redwood City, California, Us

Office Space utilization project: Summary: Measured real-time workplace utilization by developing vision-based system for human detection using deep learning to support strategy making and workspace optimization.1. Applied transfer learning by TensorFlow and Keras, introduced data augmentation and label smoothing to finetune pretrained deep learning models such as MobileNet, Mask R-CNN to detect, count people through live camera on GCP VM and boosted 10%+ mAP; Extracted numerical data of people present in each office zone.2. Assisted data scientists to transform, analyze unstructured datasets in Python, and created Tableau dashboards to explore the workplace utilization and usage.3. Prepared monthly presentations and demos to report project progress to directors and VPs; organized five knowledge share workshops for 20+ colleagues, covering Statistics, Python and Machine Learning.

1. Constructed algorithm for farm insurance pricing to help protect farmers’ profits from volatile agricultural prices.2. Built ETL tools for agricultural data and predicted agricultural prices by training ML models to support decision making on premium.3. Contributed in building an innovative pricing model in Python; simulated portfolio results; managed portfolio risk; created interactive dashboard to integrate, analyze and derive data insights in order to optimize risk.

Master Of Arts - Ma, Statistics - Data Science Track

Bachelor Of Science - Bs, Mathematics

Bachelor Of Science - Bs, Applied Mathematics