Data Science Power Bi

Harshit Arvind Barde

**************@*****.*** 313-***-**** United States

EDUCATION

Master of Science in Data Science, University of Michigan-Dearborn Dec 2024 Relevant Coursework: Artificial Intelligence, Database, Cloud Computing (GPA: 3.72/4.0) Multivariate Statistics, Bigdata Visualization and Analysis, Intelligent Systems, Deep Learning Bachelor of Engineering in CSE, Nagpur University July 2022 Relevant Coursework: Data Structure, SQL, C, System Design, Cloud Computing, Compiler Design, Operating System (GPA: 3.77/4.0) Certification:

• Python for Data Science and AI, Coursera (Python, Pandas, Matplotlib)

• Data Analysis with Python, Coursera (Seaborn, Graphs, Power BI)

• IBM Data Science: Professional Certificate

SKILLS

Programming: Python SQL R C C++ Machine Learning: Decision Trees Random Forest XG boost LSTM Databases: MySQL Teradata Big Querry Visualization: Tableau Power BI Plotly Dash Qlik Sense Big Data: Hadoop Pyspark ETL Hive Cloud: Azure Databricks GCP FireBase Vertex AI Terraform Other Technologies: MongoDB Git Apache Spark dbt Docker Airflow PROJECTS

Walmart Sales Forecasting using Python:

• Developed Walmart sales forecasting model using Python, AWS S3, Redshift, ARIMA, and XGBoost for accurate predictions.

• Utilized AWS Kinesis for data preprocessing, managing large datasets, handling missing values and outliers, and applying advanced feature engineering techniques to enhance model accuracy, accounting for peak sales and holiday effects. Graph Visualization using IMDB Data:

• Developed network visualizations using Python and Power BI on a dataset of 100,000+ IMDB movies, mapping collaboration patterns between actors and directors.

• Analyzed genre-specific collaborations with Power BI, revealing insights into how these relationships impact genre popularity and movie production.

Sterring Maneuvering Recognition:

• Led the development and deployment of LSTM-based steering models using Databricks, AWS, achieving 95% accuracy and enhancing driver safety by 40%.

• Applied LSTM algorithms in MATLAB for precise steering recognition, reducing response time to hazards by 30%. Vehicle Battery State of Charge Estimation:

• Led an AWS cloud initiative for vehicle battery state of charge estimation, achieving 99.2% precision and 0.8% MSE in %SOC prediction using ML and Neural Networks on AWS, Databricks.

EXPERIENCE

Data Analyst University of Michigan Jul 2023 – Present Patent (Ref – 2023-591): "Personalized Graduation Experience: Automated Name-Pronunciation with Real-Time Recognition":

• Faced with the challenge of automating name pronunciation during graduation ceremonies to ensure accuracy and real-time synchronization with live feeds.

• Designed and deployed a Python-based solution with AWS S3, BigQuery, and Firebase for data storage and processing, achieving 147ms latency with Flask and WebSocket. Leveraged Google Text-to-Speech API for accent variations, integrated Name Coach API for secure data retrieval, and applied Pydub for frequency-based voice normalization.

• Successfully enabled seamless, culturally responsive, real-time name pronunciation, significantly enhancing the graduation experience through precise pronunciation and synchronized visuals. VNIDS (Vehicle Network Identification System):

• Leveraged Python, AWS, and big data tools like Hadoop, Hive, Spark, and EMR to create a robust system for processing large datasets, enhancing vehicle cybersecurity with unique digital identifiers.

• Integrated ETL tool SSIS to streamline data transformation, utilizing Google BigQuery and Graph Neural Networks (GNN) for improved data storage and insights.

• Implemented AWS encryption to ensure compatibility across OEMs, presenting this innovation at automotive tech expos to showcase its real-time vehicle identification capabilities. Software Engineer Tietoevry India Pvt Ltd Jan 2022 – Dec 2022

• Designed a system to handle clickstream data from Adobe Omniture, utilizing Hadoop, Hive, Spark, and Amazon EMR, which reduced processing time by 50%, significantly improving workflow efficiency.

• Leveraged ETL tool SSIS to consolidate data integration processes, ensuring seamless extraction, transformation, and loading across diverse sources for enhanced data consistency.

• Applied AI using Python and AWS SageMaker to integrate predictive analytics and automated reporting, improving system performance and enabling accurate, data-driven insights.

PUBLICATION

"iMee: IMPLEMENTATION OF CUSTOMER-ORIENTED ERP”– PaperID: IJISRT22APR1501- Volume/Issue: Volume 7 - 2022, Issue 4 – April

Contact this candidate