UOA CS778 COURSEWORK PROTOTYPE
IdeaSense AI Coursework Prototype
A five-person University of Auckland CS778 team prototype that explored a staged AI conversation and confirmation workflow.
Project context
Problem and constraints
- The team needed a clear workflow for users to progress through an AI-assisted conversation and confirm a summary.
- The frontend had to present a multi-stage product flow while integrating with the team system context.
Personal contribution
- Implemented most of the frontend for the five-person coursework team.
- Built the project workspace and staged conversation interface.
- Implemented the summary-confirmation flow and report presentation.
Architecture and workflow
- A staged Next.js and React workspace guides the user from conversation through summary confirmation and report presentation.
- FastAPI, PostgreSQL, and AWS Bedrock describe the supporting team-system context.
Implementation decisions
- Use Next.js, React, and TypeScript for the product interface.
- Use Tailwind CSS to keep the staged UI consistent.
- Test frontend behaviour with Jest and Playwright.
Technology map
Frontend
Frontend verification
System context
Verification and evidence
- The coursework prototype was delivered by a five-person University of Auckland CS778 team.
- Frontend work covered the project workspace, staged conversation interface, summary-confirmation flow, and report presentation.
- Jest and Playwright were part of the frontend verification stack.
- The coursework deliverable provides evidence of the team prototype.
- Jest and Playwright support frontend-level verification.
Limitations and unfinished work
- This case study covers the original CS778 coursework prototype only.
- It does not claim ownership of the later independently operated IdeaSense product.
Lessons and next steps
- A staged conversation benefits from an explicit confirmation boundary.
- Team product delivery needs clear ownership across frontend and supporting services.