Project Overview
The project offers students the opportunity to delve deeper into a specific topic related to AI systems and AI infrastructure, addressing a practical pain point in the real world. The instructor will supply a list of project ideas to students. Students will formulate a problem, design and implement the system, evaluate the resulting prototype, and present a working demonstration during the final project showcase. The course will provide supporting resources as available, including coding-agent API credits and access to GPU computing through platforms such as Google Cloud, AWS, and UVA Rivanna.
Requirements
Team: Form a team of two members to start the project. Contact the instructor if you need help in finding your teammates.
Topics: You are highly suggested to choose a topic from the instructor’s project idea list. We prepared a list of big topics where each big topic is divided into multiple smaller, palatable projects that are self-contained while correlated with each other.
Proposal report: Submit a brief project proposal outlining your chosen topic, your research questions, and a preliminary outline of your design. The proposal should be 1-2 pages long.
Checkpoint presentation: Prepare a 5-minute lightning talk to present your project progress.
Final report: Write a report summarizing your findings. The report should include an introduction to your topic, a summary of your design, implementation, evaluation, and findings, and a conclusion discussing the implications of your work for serverless AI. There is no strict requirement on report length.
Project presentation: Prepare a 10-15 minute (tentative, and will be determined by class size) presentation to share your findings with the class. You should aim to make your presentation accessible to all students, regardless of their familiarity with your specific topic. You are highly encouraged to include a live (or video-recorded) demo in your final presentation.
Assessment Criteria
Your project will be assessed based on the following criteria:
- Depth of understanding: Does your project demonstrate a thorough understanding of your chosen topic?
- Quality of deliverables:
- A demonstrable MVP: How complete, robust, practical, and efficient is your system?
- A complete report: Is your report well-structured, clearly written, and properly referenced?
- Presentation & demo: Can you effectively communicate and pitch your idea and product to the class? Is the presentation engaging and the demo compelling?
Project Milestones
- Week 4: Team signup
- Week 7: Proposal report submission
- Week 10: Checkpoint presentation
- Week 15-16: Project presentation and report submission