Shubham Mishra

Experience

• 

  Penn State University

  Jan 2021 - Dec 2021

  ▪Held office hours for PSU 17 and solved doubts along with grading homework. ▪Worked with Penn State it to help solve real problems of customers around the world.▪Solved customers’ real time technological problems using online phone panel and 24/7 chat feature.▪Assisted supervisors on Research projects for upcoming technological developments in enhancing Artificial Intelligence tools for the IT Department at Penn State. ▪Held office hours for STAT 200 and solved doubts along with grading homework.

  • Learning Assistant

    Aug 2021 - Dec 2021

  • Information Technology Help Desk Support

    Jan 2021 - Dec 2021

  • Learning Assistant

    Jan 2021 - May 2021
• 

  Intelligent Direct, Inc.

  May 2021 - Aug 2021

  Data Analyst

  ▪ Utilized Python-Scikit-learn to implement machine learning models, specifically logistic regression, to extract and visualize specific data from a geographic dataset in Power BI for distribution to clients.▪ Developed a comprehensive dashboard using Python, SQL, Excel, and R to display the real-time location-based data of service requests across the United States. This automation of the process saved over 40 hours of labor- intensive work.▪ Contributed to the development and maintenance of a MySQL database by making essential modifications to its structure, leading to a 45% reduction in the risk of data corruption.▪ Conducted data cleansing and analysis on a dataset of 10,000 records utilizing Excel and Power BI pivot tables, conditional formatting, and other advanced data manipulation techniques to extract insights and improve data quality, resulting in a 20% increase in accuracy and efficiency. Show less

• 

  Croptix

  Jun 2022 - Aug 2022

  Data Analyst

  ▪ Developed a comprehensive report for the client by utilizing Python in Jupyter Notebook and Tableau to provide dynamic and real-time insights into product performance and critical business KPIs.▪ Utilized Pandas as an API to process over 100,000 time series data points, performed extensive data wrangling to ensure 98% data quality, and constructed over 20 visualizations to support exploratory data analysis and data- driven decision-making.▪ Designed and developed an internal data visualization platform for clients to analyze historic data from QGIS, resulting in enhanced informed decision-making for strategic land purchases.▪ Applied XGBoost, a machine learning algorithm, to existing regression models to detect maximal linear correlations between magnesium production and other factors, resulting in the discovery of new statistical inferences and improved data-driven decision-making.▪ Preprocessed, cleaned, and prepared 5,000 rows of data from clients using SQL and Excel to build a logistic regression model between two independent variables, resulting in an accuracy rate of 85% for accurate forecasting and enhanced understanding of key relationships between variables. Show less

• 

  Penn State University

  Jul 2022 - Oct 2022

  SIMBA SAP Research Intern

  ▪The Penn State SAP Next-Generation Program allowed me to assist with the evaluation of Tricentis Tosca tools for process automation analytics with the Penn State University SAP ERP production testing protocol.▪Focused on the 3-in-1 implementation that included SAP Business Process Management, Automation of Processes, and Digital Adoption Platform that supported 5,000 Core PSU SAP S4HANA users and 15,000 total users including Penn State faculty.▪Assisted with the implementation of scalable test management, operations, and analytics to centralize testing and orchestrate quality at speed, with visibility throughout the Penn State SAP software development lifecycle. The pilot resulted in approximately 80% of PSU SAP ERP test savings production support time.▪Collaborated with a cross-functional team of supervisors and researchers to analyze and evaluate SAP ERP testing protocols that was the foundation of the Penn State interdisciplinary research portal. This effort was the basis of PSU programs valued approximately at $1.239 Billion in Research Expenditures. Show less

• 

  Questrom School of Business, Boston University

  Jan 2024 - Jan 2025

  Research Assistant

  ▪Spearheaded RA team operations, enhancing task efficiency through the implementation of agile methodologies and problem-solving strategies, significantly reducing project turnaround times. Pioneered innovative approaches to address complex challenges, fostering a culture of creativity and solution-oriented thinking.▪Executed advanced geospatial data analysis using Folium and Google Maps API, successfully geoencoding locations lacking coordinates. Specialized in sophisticated data manipulation and transformation techniques, leading to the development of dynamic, user-friendly interactive maps with informative popups, thereby demystifying subsidiary information for non-technical clients. Show less

• 

  Spinnaker Analytics

  May 2024 - Aug 2024

  Business Intelligence Analyst

  ▪Preprocessed over 1 million online sales records for 1,142 products across 5 years by aggregating data, identifying discontinued products, and filling missing months using linear interpolation techniques to ensure data continuity.▪Developed and implemented advanced time series models including ARIMA, SARIMA, SARIMAX, Double and Triple Exponential Smoothing, and LSTM for predictive modeling of inventory needs, achieving improved accuracy in forecasts.▪Optimized forecasts by reducing prediction error to less than 30% for 78% of the products and achieving less than 10% error for 45% of the products by fine-tuning and stacking LSTM, ARIMA, and SARIMA models.▪Performed detailed statistical analysis to assess stationarity, seasonality, and trends, and used PACF/ACF plots to guide model selection, ensuring robustness and accuracy in forecasting models.▪Evaluated model performance across 341 product categories, producing comprehensive performance metrics and model comparisons, including ARIMA's 28.55% error and Triple Exponential Smoothing's 2.77% error for key product forecasts.▪Collaborated with a team of analysts and data scientists to manage the project’s deliverables using Agile methodologies, ensuring the timely development and deployment of models for strategic decision-making. Show less