From research in Germany to building with emerging tech in New York, I’ve always followed curiosity, not comfort. That curiosity led me to explore AI, product thinking, and entrepreneurship—where I get to combine strategy, data, and impact in ways that matter.

I’m a Computer Science & Business Honors student at Lehigh University. I’m currently interning in AI & Data at EY, a fellow in Cornell Tech’s AI/ML Studio, and building projects that make real problems simpler, smarter, and more human.

⚠️ Quick note: My commit activity looks quieter this year (especially in spring and summer) because my Software Engineering class used a private repo, and most of my AI/ML work was done through platforms like Canva and Jupyter notebooks — so a lot of that effort doesn’t show up here 😔

Check out my work on a machine learning project that analyzes U.S. Census data to predict whether an individual's income exceeds $50K. This was part of a larger applied machine learning course focused on data preparation, model tuning, and real-world evaluation metrics.

Tools: Python, pandas, scikit-learn
What I Did: Cleaned and one-hot encoded 32K+ rows of census data, trained and tuned Logistic Regression and Decision Tree models using GridSearchCV, and evaluated performance with accuracy, recall, and overfitting analysis.
Results: Achieved 85%+ accuracy and 58% recall, selecting the optimal model configuration by balancing generalization and interpretability.
👀 Curious? See the full project here: github.com/zaki-m-khan/census-income-prediction

Python • Java • C/C++ • SQL
Langchain • Pinecone • Azure AI (GPT-4, Document Intelligence, Speech) • Amazon Rekognition
scikit-learn • pandas • NumPy • Power BI • Databricks
React.js • Node.js • Git/GitHub • AWS (Lambda, S3, DynamoDB, Location) • Jira • Excel (VBA)

AITaskBuddy
Built for the Microsoft Innovative Challenge — a tool that helps disability job coaches automate planning and progress tracking.
Tech: Python, React, Node.js, 8+ Azure AI services (GPT-4, Speech, Doc Intelligence)

HawkFind (AWS Hackathon – 2nd Place)
Built a lost-and-found platform using Amazon Rekognition and AWS serverless tools. Cut average search time by 97% and boosted campus recovery rates.
Tech: Rekognition, Lambda, S3, DynamoDB, Location Service

AI Studio Project – Coming Soon
Currently working on a capstone through Cornell Tech’s AI Studio. Applying CRISP-DM on real-world data using deep learning, tree ensembles, and model optimization. Will be pinned once live.

• Experimenting with Retrieval-Augmented Generation (RAG) pipelines using Langchain + Pinecone
• Applying prompt engineering to benchmark and enhance chatbot accuracy
• Exploring how AI can enable equity in access, especially in education and career pathways

• Leading the Lehigh CSB Association, supporting 280+ students exploring tech and business
• Founded the Lehigh AI Club, now 250+ members strong, running practical workshops and speaker events
• Helping scale Propel2Excel, an accelerator supporting overlooked student talent across the country
• Always working to bring more voices to the table—especially in spaces where innovation meets impact

Email: zmk227@lehigh.edu
LinkedIn: linkedin.com/in/zakimkhan
GitHub: github.com/zaki-m-khan

• Azure AI Engineer Associate (AI-102)
• Databricks SQL for Data Analysts
• Propel2Excel Professional Development

Thanks for stopping by — always open to collaborating with other builders and mission-driven teams.