The best side project ideas for software engineers solve a problem you care about, teach you something new and are not a generic tutorial copy. Start with a tool you would use, ship it early and stop when it fails those tests.
This guide sits in the software engineering blog. It covers choosing an idea, checking it, fitting it around a job and knowing when to walk away.
How to choose a side project
A good side project meets two tests from CodePath’s guidance on standout side projects. It solves a specific problem that interests you, and it helps you learn something new. The same guidance adds a third filter: the project should not be a generic one lifted from a class or tutorial.
Those tests turn a long wish list into a short one. If an idea fails the first, you will struggle to keep going on tired evenings. If it fails the second, you are repeating what you already know. If it fails the third, it adds little for anyone reading your profile.
CodePath also suggests keeping a running list and adding one idea per day, regardless of how good you think it is. Judge the list later against the tests above, not at the moment you write each entry. Ideas that survive a few weeks of rereading are usually the ones worth your limited hours.
Side project ideas worth building
These ideas are not strictly ranked, because nothing in the evidence says one beats another for everyone. Read them as an ordered selection, moving from low commitment to higher commitment. Each one says what it is, who it suits and where it falls short.
A command-line tool built on the GitHub API
The roadmap.sh project list includes a beginner exercise that uses the GitHub API to fetch user activity and display it in the terminal. It suits engineers who want a small, finishable project that teaches working with a public API. The limit is that it is a listed exercise, so by the generic-project test it needs your own twist, such as a view of activity you would actually check.
A personal portfolio site
Another roadmap.sh project converts a simple HTML website into a personal portfolio. It suits anyone who needs a home for their other work and is happy to practise CSS along the way. It teaches presentation more than engineering depth, so treat it as the frame for your projects rather than the main one.
A tool that fixes your own annoyance
This is the idea that best fits CodePath’s first test, since it solves a specific problem of interest to you. It suits engineers who want something with a real user from day one, even if that user is only themselves. The risk is scope: a personal annoyance can grow into a product you never finish, so decide early what “done” means.
A project in a technology you have not used
A Stack Overflow article on side projects calls them a great opportunity to learn a new technology. Moving a small app’s data from one database to another is one way to do that; the mongodb to postgresql migration guide shows what that kind of change involves. This suits engineers whose day job keeps them in one stack. The trade-off is slower early progress, because you are learning and building at once.
A deployed service you can break and fix
A small service that real traffic, even your own, can reach teaches things a local project never shows. When it fails, the skills in how to debug production issues apply directly. It suits engineers who want operational experience outside work. The cost is ongoing upkeep, which matters when your time is limited.
Comparing the ideas
| Idea | What it mainly teaches | Watch out for | Planning use |
|---|---|---|---|
| GitHub API command-line tool | Calling an API and presenting results in a terminal | Reads as a generic exercise unless you give it a purpose | Pick it for a small, quick win |
| Personal portfolio site | Turning a plain HTML site into a presentable page | Shows front-end polish, not engineering depth | Build it to hold your other projects |
| Tool for your own problem | End-to-end ownership of something you use | Scope can grow without a definition of done | Choose it when you want the strongest story |
| New-technology project | Something new to you, which is CodePath’s second test | Slower progress while you learn | Choose it when your job keeps you in one stack |
The project descriptions come from the roadmap.sh project list, and the selection tests from CodePath.
Validating an idea before you build
Validation means finding out whether anyone else cares about the idea. CodePath notes that GitHub stars or visitor counts can serve as validation. Its author describes a 2018 proof-of-concept web app that reached 400+ GitHub stars and 5,000+ visitors.
Those signals arrive after something exists, so the cheapest check is a small, early version. The Stack Overflow article makes the same point: development is iterative, and what you make does not have to be perfect the first time. A rough proof of concept in public is a better test of interest than weeks of private polishing. The dev.to advice to pick one project, build it and share your progress follows the same path.
Side projects that help you get hired
The Stack Overflow article quotes a developer saying most technology jobs like people who have side projects. It also describes side projects as a way to put your work out where people can see it. Visibility matters most, so prefer projects that end up with a public link and a clear explanation.
GitHub is the usual place to show that work. In the 2025 Stack Overflow Developer Survey, 66.9% of respondents use public GitHub for community. GitHub scored 59.3% on desirability and 70.1% on admiration, and it was a more desirable collaboration tool than Jira that year.
- Write a proper README. CodePath advises that a GitHub link on your application should have a proper README showcasing your work, because that page is often all a reviewer reads.
- Show the proof you have. Stars or visitor counts, where you have them, back up a claim that the project mattered to someone.
- Prefer your own problem to a tutorial clone. A project with a story invites better questions than a generic one.
Balancing a side project with a full-time job
Free time rarely turns up by itself. One developer quoted by Stack Overflow said there is no such thing as spare time, so you have to make it. Developers in the same article most often named watching TV as the activity they give up for coding.
The practical advice there is to set aside a block of time every day, even a very small one, and hold yourself accountable. Small and regular is a better fit for a working week than occasional long sessions. The article also suggests that independent, exciting work is a powerful tool against burnout, so a project that energises you can help your job rather than compete with it.
When to stop a project
Stop when a project no longer passes the tests that justified starting it. If it has stopped solving a problem you care about and has stopped teaching you anything, the reasons to continue are gone. Finishing out of guilt costs hours that a better idea could use.
Three checks help you decide:
- Interest has gone. If the problem no longer matters to you, the first selection test fails.
- Learning has stalled. A project that only reuses skills you already have teaches you little, so it fails the learning test.
- The project feels like pressure. Side work is meant to fight burnout, so one that adds to it has turned against its purpose.
Stopping is not a failure. CodePath notes that not every side project needs a form of validation, and the same applies to a place on your resume. A project that taught you a technology has already done its job.
Frequently asked questions
How many project ideas are there to choose from?
The roadmap.sh project list shows 112 matches, so the shortage is rarely ideas. The harder part is choosing one that passes the tests for interest and learning.
How should I list a side project on a resume?
CodePath recommends giving each project 1-3 bullet points. It also advises keeping side projects out of the work experience section unless the work was paid. Read more in its guide to standout side projects.
Does every side project need validation?
No. CodePath says not every side project needs a form of validation, so a learning project can be worth finishing even if nobody else ever uses it.
Why do side projects help my confidence and portfolio?
A dev.to article on side projects that make you a better developer says they help you apply what you learn, build confidence and make your portfolio stand out.
