About CodeSpell
We think grading code
shouldn't consume the people
who teach it.
CodeSpell started when a CS professor spent an entire weekend grading 200 Python submissions by hand — and realised every one could have been evaluated in seconds. We built the tool we wish existed.
2025
Founded
2
Team members
2
Institutions
5
Languages supported
Our mission
Give educators their time back — without sacrificing the quality of feedback students receive.
The average CS instructor spends 6–8 hours grading code per assignment cycle. Across a semester, that's hundreds of hours that could go toward curriculum design, office hours, or research — work that only a human can do.
Meanwhile, students wait days for feedback on code they submitted hours ago. By the time they learn what went wrong, they've already built on top of the mistake.
CodeSpell closes that loop. Automated grading runs in seconds. Feedback lands while the problem is still fresh. Instructors review exceptions, not every submission.
What we believe
Principles, not slogans.
Educators first.
Every product decision starts with one question: does this save a teacher time? We build for the person grading at midnight before a 9am lecture, not for the demo.
Feedback that teaches.
A score without an explanation is just a number. CodeSpell is built to give students feedback they can act on — specific, clear, and tied to the code they actually wrote.
Honest about AI.
We use AI where it genuinely helps and don't pretend it's infallible. Instructors stay in control of their rubrics. The system surfaces problems; the human decides what matters.
Consistency as fairness.
When a rubric is applied inconsistently, students with more access to instructor attention benefit most. Automated grading is one of the most equitable tools in education.
Small team, high standards.
We'd rather ship something that works exactly right than ship fast and apologise later. Technical education is too important for sloppy tooling.
Built in public.
We share what we're building, what broke, and what we're thinking about. We think the best products come from honest conversations with the people who use them.