These projects reflect my skills across the data science and machine learning pipeline — from data preprocessing and model development to evaluation, visualization, and deployment.
Each project is thoughtfully designed, implemented, and documented on GitHub. You can explore them by clicking the images below. Every repository includes clear objectives, detailed approaches, and results that highlight both technical depth and problem-solving ability.
Together, they demonstrate not just technical proficiency, but also a strong commitment to building impactful, end-to-end solutions.
Developed an end-to-end machine learning system to predict industrial machine failures like overstrain, power surge, and tool wear.
Deployed a real-time FastAPI backend and an interactive Streamlit dashboard.
🔗 GitHub Repo
🔗 Live App
Built a scalable analytics system to track real-time retail activity using AWS.
Streamed events via Kinesis, processed them with Glue + Athena, and visualized results in QuickSight.
Created a chatbot using RAG to recommend government schemes based on age, income, and occupation.
Used Gemini embeddings + Pinecone vector DB for relevant results. Automated using n8n workflows.
Ran a randomized A/B test to measure the impact of a product change on conversion rate.
Performed SRM check (p = 0.8524) and applied Z-test/T-test to evaluate significance.
Result: +14.22% conversion uplift — recommended full rollout based on statistical confidence.
Cleaned and analyzed Chicago crime data to identify spatial and temporal patterns.
Used stratified sampling + statistical validation to ensure representativeness.
Final output: Interactive Tableau dashboard and Matplotlib visualizations.
🔗 GitHub Repo
📊 Tableau Dashboard
Designed a medallion-style data warehouse with Bronze–Silver–Gold layers.
Built ETL flows with stored procedures and applied validation checks for quality.
Visualized key KPIs in Tableau including revenue trends and customer segmentation.
Applied Modern Portfolio Theory (MPT) to optimize a portfolio of FAANG stocks.
Used Sharpe Ratio, Efficient Frontier, and Min Volatility analysis to guide allocation.
Compared strategies like equal-weighted vs optimized returns from 2020–2023.
Built an interactive aptitude performance tracker using Google Gemini Canvas to log, analyze, and visualize daily practice.
Designed a prompt-driven system to record scores across topics (Quant, DI, LR) and receive feedback on weak areas.
Experimented with zero-shot and few-shot prompting to track progress over time.