Mohammad Mahmud

The City University of New York, Hunter College
Bachelors in Computer Science

I’m a passionate software engineer and aspiring product manager who loves turning ideas into impactful AI-driven solutions!

At Headstarter AI, I build machine learning models and large language models that tackle real-world challenges. I’m dedicated to creating next-generation AI tools that empower both developers and users, making AI more accessible and intuitive.

Currently, I’m developing an AI coding assistant that integrates seamlessly into IDEs, allowing developers to collaborate with AI to enhance their code, implement new features, and streamline their workflow through an intuitive chat-based interface. My focus lies at the intersection of AI, software engineering, and product innovation, bridging the gap between cutting-edge technology and user-centric solutions.




Data Analytics Fellow at CO-OP

Cleaned and preprocessed structured and unstructured data to improve data quality for machine learning models

Conducted exploratory data analysis to identify patterns, trends, and outliers in large datasets using Python

Software Engineering Resident at Headstarter

Building Fullstack applications with Machine Learning Models and OpenAI

Creating next generation AI tools for developers and users

Software Engineering Intern

Built interactive dashboard using Streamlit and Tableau

Extracted, transformed, and loaded large datasets from multiple sources

Projects

Working on my first ML project

Customer Churn Prediction with Machine Learning

Designed and fine-tuned machine learning models to predict customer churn, achieving predictive accuracy of up to 85%.

Brain Tumor Classification with Deep Learning

Developed a deep learning model using transfer learning and a custom CNN architecture to classify brain tumors,
achieving over 95% accuracy on the test set.

Pentagram

Image generation agent. Using Hugging Face test-to-image generation model created a full stach application where user can generate an image using text input

IDE Agent

Imprelemted AI coding assistant agent like cursor on a pre-built ide. User can chat with the AI to add code or improve code.

GitHub AI Review Agent

Implemented Python parser and JavaScript parser to help AI agent to understand code for review.

CodeBase RAG

Fullstack application using AI to answer question. AI agent can read your github repository to answer questions based on your repository link