Algo Platform
Algo Platform
A course-originated project that I continued shaping into a clearer full-stack MVP, with more attention to system boundaries, local persistence, and desktop workflow.
2025
A local-first desktop coding practice platform integrating problem data, submission workflows, local persistence, API layers, and an in-app assistant.
Overview
Algo Platform began as a team course project and later became a personal redevelopment effort to make the MVP more coherent across data, API, UI, and desktop execution.
Problem
Coding practice tools often split the work across problem lists, submissions, execution, and help. For this project, the challenge was to keep those pieces connected in a local-first desktop experience rather than a browser-only prototype.
Role
End-to-end contributor across schema design, API implementation, UI integration, and testing, with emphasis on the desktop MVP path.
Context and constraints
The starting point was team coursework, so the codebase already had inherited decisions and uneven boundaries.
Desktop delivery, local persistence, and assistant behavior needed to fit together.
The project had to stay maintainable enough for continued iteration after the course submission.
What I built
A clearer data-to-API-to-UI path for problems, languages, and submissions.
A desktop delivery path using Electron rather than stopping at a web-only prototype.
Initial context-aware assistant integration inside the practice workflow.
Technical approach
Used Electron and React for desktop UI.
Kept the backend structured with Express, TSOA, and Prisma.
Used SQLite for local persistence in line with the local-first direction.
Outcomes
Extended the system beyond a course prototype into a fuller desktop MVP with structured problem and submission workflows, local SQLite persistence, and clearer frontend/backend separation and system structure.
Improved consistency between persistence, API behavior, and interface flow.
Created a practical full-stack example that complements the AI and healthcare systems work.
Reflection
The useful work here was integration: tightening boundaries, connecting layers, and making the MVP easier to continue rather than adding features for their own sake.