I am a data science postgraduate passionate about uncovering patterns, building scalable systems, and applying analytics for impactful decision-making.
- Education: Postgraduate in Data Science and Decisions — University of New South Wales (Feb 2023 – Jan 2024)
- Location: Sydney, Australia
- Interests: Data analytics, econometrics, statistics, and cloud technologies
- Contact: meetsavsani2000@gmail.com
Overview
Analyzed the impact of five renewable energy policies on:
- Growth in renewable energy adoption
- Reduction in CO₂ emissions
Compared outcomes before and after the Kyoto Protocol, using panel data for 60 countries from 2000–2020.
Data Sources
- Socioeconomic data — World Data Bank
- Energy data — International Energy Agency, Our World in Data
Data processing performed in R using dplyr and tidyr.
Methodology
- Approach: Difference-in-Differences (DiD)
- Model: Ordinary Least Squares (OLS) regression
- Controls: 10 socioeconomic variables
- Validation: Extensive robustness tests on panel microeconomic data
Key Findings
| Impact Area | Positive Policy Impact | Negative/No Impact |
|---|---|---|
| Renewable Energy Share | Feed-in Tariffs, Tradable Green Certificates (TGC), Tax Rebates | Subsidies, Auction-based Tenders |
| CO₂ Emissions Reduction | Subsidies, Auction-based Tenders, TGC | Feed-in Tariffs, Tax Rebates (no significant effect) |
Project Details
- Tech Stack: Python, R, LaTeX
- Skills: Research, documentation, panel data analysis, policy evaluation
- Full Report
Overview
Developed a full-stack system for financial data ingestion, relational storage, and AI-assisted querying with real-time visualization.
Data Sources
- Market data and financial statements from Yahoo Finance via the
yfinancePython library - Includes stock prices, income statements, balance sheets, and cash flow reports
Methodology
- ETL Pipeline:
- Extracted data using
yfinance - Transformed and cleaned with
pandas - Loaded into PostgreSQL via
SQLAlchemyORM
- Extracted data using
- Database Design:
- Normalized relational schema with indexed tables
- Designed for scalability, query performance, and data integrity
- AI Query Interface:
- Built with Streamlit
- Integrated LangChain + OpenAI API for natural-language-to-SQL translation
- Enabled dynamic queries and interactive visualizations
Project Details
- Tech Stack: Python, Pandas, SQLAlchemy, PostgreSQL, Streamlit, LangChain, OpenAI API
- Skills: ETL design, database optimization, AI integration, interactive data visualization
- Repository
- Programming: Python, R, SQL, PySpark
- Data Engineering: Apache Spark, Apache Kafka, Docker
- Cloud: Azure, AWS, GCP, Databricks
- Visualization: Tableau, Power BI, Looker
- Other Tools: LaTeX
- Microsoft certified Azure Data Fundamentals
- Microsoft certified Azure AI Fundamentals
- Indian Institute of Technology - Roorkee certified Data Science specialist
- IBM certified Data Science Foundations
University of New South Wales
Masters in Data Science and Decisions (Feb 2022 – Jan 2025)
Key Courses:
- Statistical Data Modelling
- Python Programming
- Data Visualisation
- Machine Learning
- Econometrics
- Statistical Inference
- Multivariate Analysis