ROBERT(JIADI) ZHANG

424-***-**** ady6bm@r.postjobfree.com LinkedIn: www.linkedin.com/in/jiadi-zhang-6034341b5 GitHub: https://github.com/jiadiz?tab=repositories Education

Master of Science in Business Analytics, Carnegie Mellon University – Tepper School of Business Sep 2022 - May 2023

● Selected Coursework: Data Exploration & Visualization, Optimization for Prescriptive Analytics, Modern Data Management, Machine Learning for Business Applications, Data Analytics in Finance, Marketing Analytics, Business Communications Bachelor of Arts in Statistics (incomplete,146/150 credits), University of Minnesota, Morris Aug 2021 - May 2022 Minor in Statistics & Bachelor of Arts in Psychology, University of Minnesota, Morris Aug 2016 - May 2020

● Selected Coursework: Commercial Statistical Analysis, Data Analysis, Multivariate Analysis, Hypothesis Testing, Calculus II, Probability & Stochastics, Survey Sampling, Principles of Macroeconomics, Group Communication Skills

● Programming: Python (Gurobi, Scipy, Pandas, Numpy, PyTorch, Jupyter, Plotly), R, MySQL, MongoDB, JavaScript,

● Software skills: Tableau, Power BI, MS Excel, AWS, Azure, FastAPI, DataBricks, SnowFlake, Tableau Prep Builder,

● General skills: ETL, Data Mining, Hyperparameter Tuning, SQL Queries, Feature Engineering, Model Validation, Performance Monitoring, Risk Management, Monte Carlo, LTV Estimation, Promotion Optimization, Routing Optimization, Explanatory Analysis, GitHub, Deep Learning, Unit tests, Computer Vision, Risk Analysis, Large Language Models, Prompt Engineering, Chatbot, Linear programming, SKlearn, Data Science, Data Analysis, Hyperion, QlikView, Agile PLM, SharePoint Professional Experience

Data Analyst, SolarMax Technology, Inc. Riverside, CA Sep 2021 - May 2022

● Extracted raw data from the company’s NetSuite SQL database and transformed it into analyzable datasets for dashboard developments and data analytic projects.

● Conducted A/B testing in collaboration with the marketing department to validate the company website’s renewals. Successfully identified effective new designs, leading to a 9% increase in monthly click-to-lead conversion rate in 2021 Q4.

● Developed an explainable linear regression model in R and made it accessible through Excel to enable sales representatives to estimate residential solar project time consumption, reducing customer complaints and customer service workload by providing customers with more accurate time expectations.

● Created a dashboard in Tableau to track the average time consumption of each step in solar installation projects and developed a KPI to monitor the installation team’s overall efficiency, provided the sales support manager with a consistent platform to identify inefficiencies within the workflow and examine the effect of process optimization measures.

● Conducted time-series forecasting to calculate the depletion rate of the state-rebate pool, revealing an extended life-cycle for the company’s rebate-focused sales plan. This insight prevented the company from prematurely ceasing the strategy, allowing the company to gain 40% more revenue in 2022 from the rebate-focused sales strategy that was initiated in 2020.

● Utilized anomaly detection and time-series forecasting to identify the geographic areas with the highest future demand in SoCal, optimizing the company’s branch opening strategy by providing a list of locations with the highest estimated future profitability.

● Uncovered detailed sales statistics about all major competitors and created an index to rank each of their overlap with the company’s selling domain, allowing the company to identify true competitors and adjust its pricing, product, and marketing strategies to win market share over them.

Rebate & Consumer Analyst, SolarMax Technology, Inc. Riverside, CA Oct 2020 - Sep 2021

● Leveraging Google Cloud’s geocoding API and R’s sf package to filter out customers at risk of fire-cause outages. Targeted them for housing-battery promotion by delivering the list of filtered customers to the sales team, serving as a robust lead source that led to 7% of all housing battery sales in 2022.

● Segmented customers according to their rebate eligibility criteria and collaborated with the sales team to refine targeting strategies, focusing on customers eligible for high rebates. Achieved a notable increase of 10% in housing battery sales in 2021. Project Experience

Algorithm Design for Logistic Planning, PGT TRUCKING, Inc. Pittsburgh, PA Jan 2023 - May 2023 content: https://github.com/jiadiz/Routing-optimization-with-time-management

● Developed an algorithm that uses integer programming and network flow to find the optimal matchups between hauling requests and drivers while simultaneously creating drivers’ schedules. Simulated experimentation, demonstrating that the algorithm leads to a minimum of 16% increase in weekly revenue collection and a 5% increase in drivers’ weekly hauling volume.

● Acted as the primary liaison who bridged the communication between PGT and the algorithm team by conveying PGT’s requirements, delivering weekly progress updates, and explaining technical intricacies and practical implications of the algorithm.

● Improved the visibility of PGT’s operations by building dashboards in Python and Tableau that monitor operations efficiency and drivers’ progress within their assigned hauling routes. Bitcoin Price Forecast for High-Frequency Trading, Carnegie Mellon University, Pittsburgh, PA May 2023 - Aug 2023 content: https://github.com/jiadiz/Predicting-Bitcoin-Price-Using-Neural-Network

● Conducted time-series-purposed feature engineering and deployed MLP and LSTM through Sklearn and TensorFlow in Python to predict the Bitcoin price for hourly-frequency trading, achieving an R-squared of 0.08 on capturing price change.

● Simulated hourly-frequency trading on historic data and built an auto-trading robot, achieving a 16% 2-month cost-free return. LendingClub Loan Risk Prediction API, Carnegie Mellon University, Pittsburgh, PA Jan 2023 - May 2023 content: https://github.com/jiadiz/API-for-loan-selection

● Built a RESTful API that utilizes binary classification in Python to predict the default probability of P2P loans in LendingClub.

● Utilized LDA natural language processing to feature engineer textual data to allow the API to assess risk through Short loan Purpose Statements and job titles.

● Stabilized the API’s accuracy by stacking the predictive results of two logistic regressions and two gradient boosting(LGBM & XGB) models, achieving an AUC value of over 0.90.