Arham
Shah

Data Science · ML Engineer · Python · AI/GenAI

📍 Delhi, India · VIT Vellore
Arham Shah
B.Tech CS
VIT Vellore
Python Machine Learning GeoPandas LSTM Scikit-Learn Random Forest Data Science Logistic Regression Generative AI Prompt Engineering Quantitative Finance Claude Code Linux GitHub SQL & DBMS Python Machine Learning GeoPandas LSTM Scikit-Learn Random Forest Data Science Logistic Regression Generative AI Prompt Engineering Quantitative Finance Claude Code Linux GitHub SQL & DBMS

Computer Science undergraduate at VIT Vellore with a strong focus on data science, machine learning, and AI systems. Experienced in building end-to-end ML pipelines spanning exploratory data analysis, feature engineering, statistical modeling, and model evaluation, with hands-on work in time series forecasting, classification, clustering, and geospatial analysis. Proficient in Python, Pandas, NumPy & Scikit-learn. Actively exploring generative AI and large language models.

AIR 11
LIMIT · ISI Bangalore
Gold
WorldQuant IQC 2026
10+
Certifications Earned
What I've Built
MedhaVIT
Built an AI-powered quiz platform using Groq API to generate questions aligned to Bloom's Taxonomy cognitive levels. Engineered dual-role dashboards for teachers (quiz creation, analytics, score distributions) and students (quiz attempts, detailed analysis). Implemented anti-cheat measures including DevTools blocking and copy prevention. Integrated Supabase authentication and PostgreSQL backend , deployed on Netlify.

Link: https://medhavit.netlify.app
React Groq AI Supabase PostgreSQL Recharts
Telecom Churn Prediction
Developed customer churn prediction models on 7,043 telecom records using Logistic Regression and Random Forest, achieving 79.9% accuracy and 0.8378 ROC-AUC. Analyzed 20 features to identify key churn drivers including tenure, monthly charges, and contract type, supporting targeted retention strategies.
Random Forest Logistic Regression Scikit-Learn SMOTE EDA
Offline Reverse Geocoding — India
Built a fully offline reverse geocoding pipeline using Python, pandas, GeoPandas, and Shapely with 6+ GIS datasets to map coordinates into 6 fields including State, District, PIN Code, and Capitals. Automated spatial joins, validation, and Excel processing with 100% offline execution and zero API dependency for scalable location intelligence across India.
GeoPandas Spatial Join Pyogrio Python GIS
Teded — C Text Editor
Built a terminal text editor in C for Linux using raw terminal mode and ANSI escape sequences, implementing multi-line editing, scrolling, file I/O, and status rendering. Engineered 6 modular source files with zero external libraries for low-level systems programming.
C Systems Programming termios Linux Makefile
Academic Background
B.Tech — Computer Science & Engineering
Vellore Institute of Technology · 2025 – 2029
Coursework
Data Structures & Algorithms Operating Systems Computation Structures Object-Oriented Programming
Active member of the math and competitive programming communities on campus.
Class XII - 87.4%
P. D. Model Sr. Sec. School · 2024 – 2025
Class X - 92.8%
Narayana E-Techno School · 2022 – 2023