Open to opportunities · Pune, India

Anas Quazi Data Science & Machine Learning
2nd Year CSE · DY Patil International University

I build end-to-end ML systems — from raw data collection to deployed pipelines. Analyst by instinct, engineer by practice. I don't just run models, I understand why they work.

View Projects Get in Touch
CGPA 9.2
LeetCode Problems 100+
ML Projects 4
Engineered Features 43
GNN vs XGBoost MAPE 7% vs 12%
what I work with

Technical Stack

Languages
Python C++ SQL Java C
Data Science
Pandas NumPy Matplotlib Seaborn EDA Feature Engineering
Machine Learning
Scikit-learn PyTorch LightGBM XGBoost SHAP SMOTE
Geospatial & Graph
GeoPandas Folium NetworkX Graph Neural Networks
Databases
PostgreSQL MySQL Window Functions CTEs
Tools & Deployment
Flask Streamlit Git GitHub VS Code Jupyter
things I've built

Projects

01
Python · Pandas · Scikit-learn · Flask · Matplotlib · Seaborn
Petro-Analyst — India Fuel Price Prediction System

Predicting LPG & fuel price impact of Middle East geopolitical conflicts on Indian households. A problem no existing system was solving.

16 years of data collected from multiple sources — GPR index, Brent crude, USD/INR, LPG & fuel prices
43 engineered features — lag, rolling averages, % changes, interaction terms
KNN similarity analysis to surface historical parallels to current market conditions
Deployed interactive risk dashboard via Flask — real-time classification & 3-month forecast
Ongoing 02
Python · PyTorch · GeoPandas · XGBoost · NetworkX · Streamlit
Geospatial Property Valuation — GNN-based Real Estate Pricing

Proving that a house's price is fundamentally spatial — not just tabular. Traditional AVMs treat each property in isolation. This doesn't.

Built KNN graph where every house is a node, edges connect spatial neighbors
Graph Neural Network with attention mechanism aggregates neighborhood pricing signals
Achieved ~7% MAPE vs ~12% XGBoost baseline — spatial context wins
Interactive Streamlit dashboard with spatial price heatmaps via Folium
Completed 03
Python · LightGBM · SMOTE · SHAP · Scikit-learn · Streamlit
IoT Predictive Maintenance — Industrial Failure Prediction

Most ML maintenance systems fail in production because they ignore external context. This one doesn't.

Contextual Data Fusion — IoT sensor telemetry fused with weather & factory load signals
LightGBM with SMOTE inside cross-validation folds for imbalanced classification
SHAP explainability — which sensor triggered each failure prediction and why
Alert dashboard showing which machines will fail in the next 7 days
Completed 04
PostgreSQL · SQL · Window Functions · CTEs · Database Design
F1 Business Analysis — Multi-Billion Dollar BI System

Formula 1 is a $3B+ business ecosystem. This models its entire financial structure in SQL.

14-table normalized schema — driver salaries, sponsorships, engine contracts, factory expenses
Complex analytical queries — net P&L per team, partner networks, contract expiry tracking
Window functions for ranking, rolling aggregations, business intelligence at scale
Completed
where I've worked

Experience

Data Science & ML Intern
Infotact Solutions · Remote
2026
Built 2 end-to-end ML projects — IoT Predictive Maintenance & Geospatial Property Valuation
Worked with real industrial datasets, geospatial data, graph construction and GNN modeling
Applied professional DS workflow — semantic commits, branch strategy, Kanban tracking
Rural Data Analysis Intern
Sanjivani NGO · Dondaicha, Pune
May – Jun 2025
Scraped and structured public data to identify healthcare, education & infrastructure gaps
Built analysis pipeline using Python & Pandas on rural town data
Generated visualizations and insights for social impact reporting
recognition & milestones

Achievements

9.2
CGPA
Top 20 in College · DY Patil International University
100+
LeetCode
Problems solved · DSA in C++ · Ongoing
4
ML Projects
End-to-end pipelines · 2 deployed · GNN & time series
2
Hackathons
DSA & Competitive Programming · 2nd Year
let's connect

Get In Touch

Open to internships, collaborations, or a good conversation about data, ML systems, or sports analytics.

anasqazi869@gmail.com
GitHub LinkedIn Twitter LeetCode