Undergraduate Researcher in GenAI & Computer Vision
Computer Science, Drexel University (BS ’27)
📍 Philadelphia, PA
🌐 Website: https://www.arijichakma.com
📧 Email: ac4393@drexel.edu
I’m a Junior Computer Science student at Drexel University with research interests in 3D Computer Vision, Controllable Image Generation, Applied LLMs, and Human–AI Interaction (HAI). I’m a long-time developer with experience across the Full Stack, as well as in IoT (Arduino & Raspberry Pi), agentic workflows, and Model Context Protocols (MCPs).
Broadly, my research interests include:
- 🎨 Controllable image generation
- 🧊 3D generation, understanding, and reconstruction
- 🤖 Applied large language models and agentic workflows
- 🧠 Human–AI interaction (HAI)
Advisor: Prof. Feng Liu
📅 April 2025 – Present
- Conducting research on controllable image generation and applied multimodal AI.
📅 April 2026 – September 2026 (Remote)
📅 April 2024 – September 2024
- Designed and automated cloud-native workflows within the Cloud & AI team using AWS services.
- Built scalable ETL pipelines and internal tooling used by thousands of users across the organization.
Focus Areas:
Diffusion models, vision–language models, multimodal learning, controllable image generation
My projects are available at:
🌐 https://www.arijitchakma.com/#/projects
- 🥇 Best Social Hack — DragonHacks 2025
- 🎓 Breakthrough AI Fellowship — Cornell Tech (2025)
- ⚛️ Best Beginner Quantum Hack — BitCamp 2025
- 🥇 Overall 1st Place — PennApps 2024
- 🥈 Overall 2nd Place — HopHacks 2024
- 💡 Bloomberg Philanthropic Hack Winner — HopHacks 2024
- 🎨 Best Multimedia Track — DragonHacks 2024
- 📚 Dean’s List — Drexel University (2023–Present)
- 🌍 Drexel Global Scholarship (2022)
Course Assistant
- CS383: Machine Learning (Winter 2026)
- CS380: Introduction to Artificial Intelligence (Fall 2025)
I am actively seeking research collaborations, internships, and graduate research opportunities in computer vision, generative modeling, and multimodal AI.
Always open to collaborations, research, and building meaningful software.