Mykyta Lepikash was a teaching assistant for INFO 103, an introductory course at the University of Washington. With about 50 students in the class, most weeks meant roughly 100 submissions to grade: Python notebooks, handwritten work, and everything in between.
Before Ethyra, that work took about eight hours a week. It wasn’t just the time. It was racing through papers late at night, knowing some students were getting thinner feedback than others simply because there was never enough of him to go around.
He was skeptical at first. Another tool promising to help with grading can sound like a shortcut that lowers the bar. What convinced him wasn’t the pitch. It was the output. The feedback Ethyra produced was clearer and more useful than what he could write when he was rushing. It stayed grounded in the work in front of him, criterion by criterion.
He still reviewed everything. Ethyra handled the first pass; he kept the judgment. The difference was speed and consistency. What used to take eight hours now takes about one.
“I was shocked at the depth of the feedback. It sounds like something I would write, but I could never have tailored something that detailed to every student on my own. And my students love it.”
Mykyta Lepikash
What students noticed
Students started talking about how good his feedback was. From their side, it looked like he was an especially strong TA: someone who actually read their work and had something useful to say. He did read it. He just wasn’t spending an entire day rewriting the same comments over and over.
For Mykyta, Ethyra didn’t replace being a TA. It made the part of the job that matters, clear and fair feedback at scale, actually possible.