StudySouq maths website
Budget: $30 – $250 USD
The platform is already live in staging with an AI pipeline that produces unlimited, topic-specific maths questions, but it still needs a final engineering pass before launch.
What I need you to focus on is the AI-generated maths questions themselves—ensuring they are accurate, varied, more efficient and clearly mapped to each learning objective—while also closing out the remaining LaTeX and skill and progress-tracking tickets with helping visualize it more for the student.
LaTeX fixes
• Remove instances of incorrect mathematical notation.
• Correct poorly aligned equations so they centre properly inside our React components.
• Eliminate inconsistent formatting across browsers (currently using MathJax).
Skill-tracking refinement
• Implement more detailed tracking of student performance at the question, skill, and topic levels.
• Add better visualization of progress (charts are in place but need polish and edge-case handling).
• Improve the system’s enhanced adaptability based on skill level so question difficulty auto-adjusts in real time (currently uses an IRT based system)
AI question generation
You will work within an existing pipeline that uses multiple models such as GPT-4 and other models for quality check and a custom tagging layer. Optimise prompt templates, add unit tests around output validation, and tighten the question-to-answer-explanation cycle so learners see consistent, error-free content.
Tech stack in play
React (Vite-based setup)
Node.js + Express
MongoDB (Mongoose)
LaTeX / MathJax (Math rendering & formatting)
AI integration (OpenAI / custom math question generation)
Acceptance criteria
1. All LaTeX displays flawlessly across Chrome, Firefox, Safari, and Edge.
2. Question-generation pass rate ≥ 99 % on automated validation tests.
3. Progress dashboard updates on every submission with no perceptible lag (<200 ms).
4. New tracking tables and charts deploy behind a feature flag with zero downtime.
I can provide full repository access, sample data, and current failing test cases on day one. Let’s get this over the finish line and into the hands of teachers and students.
What I need you to focus on is the AI-generated maths questions themselves—ensuring they are accurate, varied, more efficient and clearly mapped to each learning objective—while also closing out the remaining LaTeX and skill and progress-tracking tickets with helping visualize it more for the student.
LaTeX fixes
• Remove instances of incorrect mathematical notation.
• Correct poorly aligned equations so they centre properly inside our React components.
• Eliminate inconsistent formatting across browsers (currently using MathJax).
Skill-tracking refinement
• Implement more detailed tracking of student performance at the question, skill, and topic levels.
• Add better visualization of progress (charts are in place but need polish and edge-case handling).
• Improve the system’s enhanced adaptability based on skill level so question difficulty auto-adjusts in real time (currently uses an IRT based system)
AI question generation
You will work within an existing pipeline that uses multiple models such as GPT-4 and other models for quality check and a custom tagging layer. Optimise prompt templates, add unit tests around output validation, and tighten the question-to-answer-explanation cycle so learners see consistent, error-free content.
Tech stack in play
React (Vite-based setup)
Node.js + Express
MongoDB (Mongoose)
LaTeX / MathJax (Math rendering & formatting)
AI integration (OpenAI / custom math question generation)
Acceptance criteria
1. All LaTeX displays flawlessly across Chrome, Firefox, Safari, and Edge.
2. Question-generation pass rate ≥ 99 % on automated validation tests.
3. Progress dashboard updates on every submission with no perceptible lag (<200 ms).
4. New tracking tables and charts deploy behind a feature flag with zero downtime.
I can provide full repository access, sample data, and current failing test cases on day one. Let’s get this over the finish line and into the hands of teachers and students.
Related categories:
Matlab and Mathematica
LaTeX
Mathematics
Node.js
Redis
Docker
MongoDB
Next.js
AI Model Development
AI Development