DevPath is an open-source web application that helps developers — especially beginners — find meaningful coding projects to build. Instead of searching through generic lists, a developer describes their situation in four inputs and DevPath returns the three best-matched projects, complete with everything needed to start immediately.
Many beginner and intermediate developers know they should be building projects to grow their skills, but they face two blockers:
- "What should I build?" — Choosing a project that is too easy wastes time; too hard and they give up.
- "Where do I start?" — Even a well-chosen project idea is overwhelming without a clear first step.
DevPath addresses both problems in a single tool.
- The user opens DevPath and enters the skills they already know (e.g. Python, HTML, JavaScript).
- They select three preferences: experience level, area of interest, and how much time they can commit.
- DevPath scores every project in its dataset against those inputs and returns the top three matches.
- Each match links to a full detail page showing: description, feature list, tech stack, a visual roadmap, learning resources, and a downloadable starter code template.
- Beginner programmers who have learned the basics and want to apply them
- Intermediate developers exploring a new language or interest area
- Open-source program participants (GSSoC, Hacktoberfest) looking for a contribution-ready project to work on
- Coding bootcamp students who need portfolio projects
Simple over clever. The recommendation engine is intentionally rule-based rather than ML-based. This keeps the code readable, testable, and accessible to contributors of all experience levels.
Modular over monolithic. Logic is split into utils/ modules so each
concern (data loading, scoring, file serving) can be understood and tested
independently.
Documented over assumed. Every function has a docstring explaining its inputs, outputs, and purpose.
All project data lives in data/projects.json. Each project is a self-contained
JSON object describing everything a developer needs: skills required, level,
interest area, time commitment, a description, features, tech stack, roadmap
steps, learning resources, and the path to its starter code file.
The dataset is deliberately small (7 projects) to keep it easy for contributors to add their own entries without needing to understand the full codebase.
Each project has a corresponding starter template in starter_code/. These
files are intentionally incomplete — they scaffold the structure and define
the function signatures, but leave the implementation as a learning exercise.
This is the "blank page problem" solution: the developer has a working skeleton
to read, modify, and extend.
DevPath is designed to be a learning ground for open-source contribution. The codebase is deliberately kept simple, the issues are labelled by difficulty, and the documentation walks through everything from setup to pull requests. It is suitable as a mentored project for programs like GSSoC, MLH, or Hacktoberfest.