Student Blogs#
Capstone students often learn the most from the messy middle of the project: negotiating scope, finding data, validating models, communicating uncertainty, and deciding what belongs in a final client deliverable. This section collects student reflections so future teams can learn from prior projects.
LLMs and Risk Management
Lessons with LLMs
A reflection on using large language models for risk scoring, relevance classification, and human-in-the-loop analysis.
Data Sourcing
How to Build a Data Pipeline When the Client Gives You Nothing
A practical guide to finding, evaluating, and integrating real-world datasets when no one hands you a clean source.
Responsible AI
Why Isn't Accessibility AI Detection Just Another Object Detection Task?
An argument for trust, conservative evaluation, and human oversight in accessibility-focused computer vision systems.
Multi-Agent Systems
How Do You Design a Multi-Agent AI System That Actually Converges?
A technical reflection on why supervisor architectures often outperform unconstrained agent swarms.