$ whoami

Ruben Saucedo...

Senior fullstack engineer building performant AI agent platforms.

role:Senior Fullstack Engineer @ Microsoft
focus:ai agents · performance · frontend
scale:100M+ daily users
side:ketzalcode.com/trainer (opens in a new tab)
$ ls ./work

Selected work

$ cat ./about.md

About me

I'm Ruben. I build at Microsoft Teams — five years and counting — on products used by 100M+ people every day. Came up through UPIITA in Mexico City, did an exchange in the Basque Country, started my career at Wizeline shipping production React for international clients, and joined Microsoft in 2020.

The work I'm proudest of has a common thread: making things fast, accessible, and resilient at a scale where small wins compound. I shipped a 30–40% performance gain on a surface used by millions, redesigned the Teams Admin App end-to-end (now the standard UI, 4.1★), and served as the team's Performance and Accessibility Champion.

Lately I'm leading our AI agent platform — building the orchestration layer, agent protocols, and toolchains that let production AI workflows actually work at enterprise scale. The frontend foundation is still load-bearing: state machines that don't lie, telemetry you can trust, UX that respects users when models misbehave.

On the side I run KetzalCode, an AI trainer platform I architected from scratch. In 2018 I co-invented an on-device machine learning system for Mexican Sign Language translation — it's a national patent.

$ tail -f ./now.log

Now

updated: 2026-05-12

Shipping a platform that grounds AI agents against governed customer data. Owning the fundamentals — performance, telemetry, resilience — that make agentic products reliable in production. Mentoring teammates and authoring design specs.

On the side, building features into KetzalCode.

Currently reading: agent eval methodologies, anything by Simon Willison, and a lot of production AI agent code.

$ contact --all

Contact

github:
github.com/RubenSaucedo
resume:
/resume

Open to interesting work. Reach out via GitHub.