Mandar Mhatre

Research Assistant

Indiana University•  June 2024 - Present•  Bloomington, IN

Engineered an ELT pipeline using DBT, leveraging Airflow DAGs for orchestration, creating a centralized data warehouse in Snowflake, centralizing data that enabled efficient querying and reduced analysis time by 60%. Implemented data quality checks and automated data cleaning processes using DBT data tests and macros, reducing data preparation time by 40% and improving overall data accuracy by 25%. Implemented A/B testing, leveraging Google analytics, boosting program efficiency by 15% based on KPIs including test scores, knowledge retention rates, and module completion times. Generated monthly reports on volunteer performance using Power BI and DAX, facilitating data-driven decision-making for program improvements that resulted in a 20% increase in average test scores.

Volunteer Data Analyst

Energy Justice Lab•  August 2023 - May 2024•  Bloomington, IN

Conducted comprehensive outage analysis using electric and gas service providers’ data, revealing critical insights into incident frequency, average downtime, and Mean Time to Repair (MTTR). Designed and deployed a data web application using Plotly Dash, transforming utility service area data into interactive visualizations that facilitate detailed outage analysis and empower residents with data-driven service provider selection. Built an integrated data models using python, combining Geospatial data with outage data of utility providers across 35 states, automating manual mapping processes, saving 20 hours monthly. Improved dashboard loading time by using Mapshaper to reduce polygon complexity resulting in 50% faster visualizations. Created a data dictionary and reference reports, leveraging tableau and Excel functions such as VLOOKUP and pivot tables, making data more understandable for the team, resulting in 30% faster task completion.

Software Engineer

GEP•  June 2021 - December 2021•  Mumbai, India

Automated inbound requisition mapping to approval workflows by building an ETL pipeline using C# and SQL, extracting data from REST APIs and loading into a SQL Server database, resulting in 60% faster purchase order creation. Developed API endpoints to send spend data from the B2B e-procurement platform, facilitating AI-based spend analysis. Created stored procedures to dynamically filter product data based on user-selected criteria, reducing load times by 50%. Leveraged Azure DevOps for CI/CD pipelines, deploying 35 critical releases across environments with a 100% success rate. Collaborated with a cross-functional team under an strict Agile (scrum) environment leveraging JIRA and Confluence ensuring productivity and timely completion of milestones.