Building With Purpose: My Summer at JGM Innovation

# 20 Hours to Minutes: What I Built at JGM This Summer

By Sebastian Giraldo

Going into my summer internship at JGM Innovation, I expected to spend most of my time

thinking like a data scientist: cleaning data, running models, looking for patterns. What I actually spent the summer doing was building tools that real people at JGM would open up and use every single day, and figuring out, project by project, what it actually takes for something you build to be trusted enough to run on.

My first real project started with a problem Waldo, JGM's founder, was dealing with every hiring cycle: bringing on interns meant writing and sending interview invites, offers, declines, and onboarding emails by hand, one at a time, double-checking names and templates along the way. I built a tool that could take a whole batch of candidates at once, greet each person by their correct first name, and prepare the right Gmail draft automatically depending on where they were in the hiring process. It still leaves the final send to a human, but the tedious part disappears.

We ran it through several live batches with real JGM email addresses before it became

something the hiring side of the team could depend on. What used to be a process that could eat up over 20 hours now takes a fraction of that, and hearing that directly from Waldo, who now uses it as part of his own workflow, is what made me realize I hadn't just built a script. I'd built something JGM could actually run on.

That project led somewhere I didn't expect: directly into JGM's partner school network in Peru. I built a tool that takes Spanish-language announcements from the school and turns them into polished, ready-to-share communications, styled to the school's own brand colors and built around Peru's calendar of civic events. Getting it right meant more than formatting text nicely. It meant learning what those civic dates actually mean to the community they were written for, and making sure the tool respected that instead of flattening it into something generic. The moment that stuck with me most wasn't in a meeting or a code review. It was hearing that the teachers at

the school were actually using it, and that they appreciated how simple and efficient it made

something that used to take real time and effort. That's a small thing to type out, but it's the

reason I wanted to work somewhere like JGM in the first place: to build something that quietly makes someone else's day easier, thousands of miles away, without them ever needing to think about the tool behind it.

Looking back, both projects were really the same lesson taught twice. JGM doesn't have the

headcount of a large company, so instead it finds the places where good intentions are being slowed down by manual, repetitive work, and builds something smart enough to close that gap. That's what innovation actually looks like here. It isn't a word on a mission statement, it's a deliberate choice to use technology so that more of the organization's real energy goes toward students, families, and schools instead of getting lost in overhead. I got to see that philosophy up close because the team trusted me with real problems instead of busywork, and because I had Waldo Guevara, Pranika Kalluru, and fellow interns like Ismael Hernandez and Zahyra Otero to build alongside.

I came in thinking this internship would be a line on my resume that said "data science." I'm

leaving with a much clearer sense of what it means to build something someone actually trusts, and I have this summer, and this team, to thank for that.