Interest in Contributing to sympy.solvers for GSoC 2026 – Math-Focused Python Background

[Raj Harsh]

Dear SymPy community

My name is Harsh Raj (you can call me Raj). I'm a pre-university student from Singapore,  starting university in August this year to study Computer Science.

I have solid experience with Python (including building structured LLM agents using Google's Agent Development Kit) as well as . I've been working on a startup project creating student helper agents that assists in math problems like a teacher would— this involves symbolic solving, step-by-step evaluation, rubric-based scoring , and handling edge cases in equations. I've used SymPy extensively in these agents for solvers and evaluators, so I'm familiar with solveset, solve, dsolve, and common pain points (like tricky non-linear or conditional solutions). I would like to add that I have never worked in any open source project and all my experience is from personal project or online courses.

I'm very excited about contributing to the sympy.solvers module for GSoC 2026 — areas like improving nonlinsolve, integrating helper solvers into solveset, handling assumptions better, or ODE-related enhancements 

I've already:
- Forked and set up the SymPy repo locally
- Run tests and played with some equations  

I'd love guidance on how to get started meaningfully in solvers:
- Are there specific good-first-issues or small enhancements in sympy.solvers (or submodules like ode.py, solveset.py) that are high-priority right now for a beginner contributor?
- Which open ideas or recent threads (e.g., from the GSoC 2026 discussions on Lambert, trig failures, or integrating helpers) would be realistic scopes for a first-timer aiming for GSoC?
- Would any mentor be open to discussing a potential proposal focused on solvers reliability for educational use cases?

I'm eager to contribute small fixes/PRs this week to demonstrate interest and learn the codebase better. Happy to hop on any chat to discuss what direction I should be heading into. 

Thank you for your time and for maintaining such an amazing library — it's already helping my AI-for-education projects a lot!

Best regards,  
Harsh Raj  
(Singapore)