Data Science · Python · Applied ML · Data Analytics

_{B.Sc. Data Science · University of Mumbai · 2026}

I build practical Python applications and applied machine-learning systems, with a growing focus on data analytics and privacy-preserving ML.

My work is centered on building, testing, measuring, and documenting useful systems rather than collecting tools or decorative metrics.

Current direction: Python development · SQL · Excel · Tableau · Applied Machine Learning · Data Analytics

A privacy-preserving federated-learning system for loan-default prediction using FedAvg, with centralized baselines and IID / Non-IID evaluation.

• Result: ~76.5% accuracy after 15 communication rounds under IID conditions.
• Baseline: 78.2% accuracy for the strongest centralized Neural Network baseline.
• Stack: Python · TensorFlow · scikit-learn

Repository →

A local-first Python application and CLI for saving, organizing, summarizing, studying, and exporting YouTube learning material.

• Test evidence: 233 passed tests across 17 test modules.
• Reproduce: py -3.11 -m pytest tests/ -v
• Stack: Python · Streamlit · CLI · pytest

Repository →

A Python CLI + Streamlit assistant for tasks, habits, and journaling with multi-provider LLM support.

• Interfaces: Rich CLI + Streamlit web demo.
• Verification: Task CRUD and habit check-in flows documented as locally tested end-to-end.
• Stack: Python · Streamlit · LLM APIs

Repository → · Live demo →

A FastAPI + SQLite personal-finance application with a web dashboard and report/export workflows.

• Live proof: Standalone GitHub Pages demo.
• Stack: Python · FastAPI · SQLite

Repository → · Live demo →

Metrics are included only where they are tied to a specific project or verified repository evidence.

Python & development
Python · CLI applications · Streamlit · FastAPI · automation · application logic

Data & analytics
SQL · Pandas · NumPy · data preparation · EDA · visualization · Excel · Tableau

Machine learning
scikit-learn · TensorFlow · model evaluation · centralized baselines · Federated Learning / FedAvg

Applied AI
LLM APIs · multi-provider integrations · AI-assisted applications

Engineering
Git · GitHub · testing · documentation · reproducibility

These are working areas, not claims of expert-level proficiency.

learning:
  - Python
  - SQL
  - Excel
  - Tableau
  - Data Analytics
  - Applied Machine Learning

building:
  - Python applications
  - Portfolio site (HTML/CSS/JS)

exploring:
  - Privacy-preserving ML
  - Applied AI / LLM integration

B.Sc. Data Science — University of Mumbai · 2026

Academic focus includes Python, SQL, data analytics and visualization, Machine Learning, AI, Pandas, NumPy, Matplotlib, and web fundamentals.

• Evidence over decoration — show the project, result, test, demo, or reproducible command.
• Practical systems — build things that solve a concrete problem.
• Continuous learning — strengthen fundamentals while shipping projects.
• Clear documentation — make work understandable and reproducible.

_{Build with evidence. Learn in public. Ship useful systems.}