Yani.

I build things. Mostly web apps, lately with a lot of AI under the hood. Stuff like UniSort (a university recommender that's pulled in 35,000+ visitors), AGILA (a campus cat management app), and Benkyou (a Japanese study PWA I made because I was getting tired of Anki).

I'm also a 2nd-year Computer Science student at Ateneo de Manila, going into 3rd year this June, specializing in Data Science and Analytics. I'll be honest, EIF was my first real internship-type thing. School projects are one thing. Building something that an actual CEO is going to test against his own writing is another. I came in still figuring a lot of stuff out, and I'm leaving with a much clearer sense of what production AI work actually looks like.

On the side, I'm a Developer at CompSAt and a Back-end Developer at GDGoC-Loyola. Outside of code, photography and badminton.

What I did at Eskwelabs

I was part of Cohort 9 of the Eskwelabs Innovation Fellowship, AI Solution Development track. I joined because I wanted real AI and product experience outside of school. Not another class project that lives and dies on a single grade. Something that ships, with stakeholders, with users.

I worked on two projects during the fellowship. Different surfaces, same idea: AI tools built for actual Eskwelabs internal use. The Slide Deck Generator helps Learning Experience Designers build instructor decks faster. The Thought Leader Drafter helps the CEO write articles in his own voice. Different problems, but both are about taking the parts of someone's job that drain hours and seeing how far AI can take them, without producing the kind of generic output that AI tools usually spit out.

The biggest gap between this and school was the production-ready bar. School projects can be impressive in a demo and fall apart five minutes later. These tools needed to hold up. They needed to handle weird inputs, fail gracefully, and give outputs that someone would actually use, not just clap at.