Peter Du

Most recently: defining the measurement & experimentation strategy of AI features within the Spring Health member experience. Overall: solving various ambiguous data science problems impacting millions of members - such as optimizing care retention, minimizing no show behavior, and driving low friction content consumption - with the goal of eliminating barriers to mental healthcare for everyone.

I worked on 2 high-impact teams in my time at Meta. A brief description of highlights: Facebook Messenger: - Developed and implemented the prioritization and value framework for 20+ Messenger features in scope for end-to-end encryption, generating an actionable feature roadmap for product and engineering teams - Drove innovation roadmap of Messenger via feature deep-dives and foundational data research (utilizing SQL, Python, and Jupyter notebooks) Ads Core ML: - Utilized quantitative modeling and experimentation to improve ad ranking and increase revenue, resulting in an average of $200K+ daily revenue saved due to ranking miscalibrations - Defined metric guardrails and alerting setup for specific signal-loss traffic segments, ensuring high-quality ranking model output - Implemented solutions for ads teams to map relationships between prediction model quality metrics and revenue fluctuations

• Influence key product decisions by constructing data driven decision frameworks • Proactively drive alignment on the most impactful user problems and solutions • Develop a rigorous and scalable experimentation process for all in-app experiments • Measure, evaluate, and present data learnings from product features through concise storytelling • Lead the foundational work for comprehensive analytics tracking within the app • Design and build ETL tables to boost efficiency across the data analytics & data science teams • Collaborate with all stakeholders - product, business development, business operations, marketing, customer success - curating relevant data points and stories that impact their short term and long term strategies

Key Responsibilities: • Automate monthly executive reports on the Jupyter notebook platform through Python packages such as pandas, numpy, and sqlalchemy • Extract and interpret user data sourced from in-house tracking, stored in Amazon Redshift databases, through the Redshift SQL language (based on PostgreSQL) • Provide key A/B testing analytics and statistical significance calculations (sample size, p-values, minimum detectable effect) to potential feature implementations • Leverage dashboarding and visualization tools such as Periscope Data, Amplitude, and Google Analytics to communicate data driven ideas company-wide • Meet deadlines of analytics requests from various departments such as sales, marketing, and executives

• Identified and mastered patterns in drink recipes which led to significant improvement in service speed along with increased responsibilities • Served as one of the few baristas who supported supervisor's financial management responsibilities such as sales charts and food counts