Build LLD Practice Platform

I want to bring DesignLens to life—a web platform where learners practice Low-Level Design, upload their code plus short design notes, and automatically receive structured, seven-dimension feedback generated by a hybrid deterministic rules engine supplemented with AI scoring.

The backend will be written in Node.js. The evaluation flow should accept solutions that we currently tag internally as “Backend” language submissions; I will rely on you to help formalise how we map that to specific runtimes (today we anticipate Java, Python and C++ sandboxes, but your guidance is welcome).

Learners should be able to sign in, create a new problem attempt, enter or attach their solution, and submit. My earlier answer of “oops” on the submission interface reflects that I am still open to the best UX approach—whether an in-browser Monaco-style editor, file upload, or a clean paste-in form, whichever you can justify for speed, reliability and ease of sandboxing.

Deliverables
• Node.js backend (Express or similar) with REST/GraphQL endpoints
• Secure code-execution microservice(s) isolated per language/runtime
• Deterministic rules engine feeding an AI model (OpenAI or equivalent) that returns scored feedback across seven design dimensions
• Minimal front-end stub (React preferred) to exercise the workflow end-to-end
• Documentation covering setup, local run, and how to extend languages or feedback dimensions

Acceptance criteria: end-to-end demo where a user submits a sample solution and receives all seven feedback scores within reasonable latency; code repo passes clean install and test on a fresh machine.

If this sounds exciting, let’s discuss your approach and timeline.