In our recent ACM chapter meeting, we had the pleasure of hosting Elijah, a software engineer at Automattic’s applied AI division, for an insightful tech talk titled “How to Write Code that Matters”. Moving beyond theoretical computer science concepts like perfect algorithms and Big O notation, Elijah shared practical advice on creating software that has a tangible, real-world impact.
The Origin of Harper
Elijah’s journey to writing impactful code began with a relatable college struggle: trying to keep up with typing lecture notes. He found that his rapid typing inevitably led to spelling and grammatical errors, but existing solutions didn’t fit his workflow. Grammarly was expensive, slow, and raised data privacy concerns, while open-source alternatives like LanguageTool were too bulky and resource-intensive.
His solution? He built his own tool. This led to the creation of Harper, an open-source, lightning-fast, and entirely private grammar checker that was eventually acquired by Automattic.
Three Steps to Writing Code That Matters
Drawing from his experience developing Harper, Elijah outlined three key takeaways for students and aspiring engineers:
Build Something You Would Actually Use If you don’t want to pay for a tool or find existing tools lacking, try building an open-source alternative yourself. Even if you end up being the only user, you will have created something tangibly productive for your own daily life.
Pick a Really Hard Problem Language processing is an incredibly complex beast, involving lexing, fuzzy parsing, and building extensive rule engines. By choosing to tackle a highly difficult problem, you guarantee that you will learn a lot. Even if the project doesn’t become a massive success, proving you can solve complex problems is an incredible asset for your resume.
Share It With the World Quickly Find people who share your specific use case and get your project into their hands as soon as possible. Don’t let the fear of imperfect code hold you back—what matters most to early users is simply that the software works.
Elijah’s talk was a fantastic reminder that the most meaningful software often starts by simply trying to solve our own everyday problems. A huge thank you to Elijah for sharing his story and advice with our chapter!