Software engineer · Seattle

MonicaGonsalves

Backend systems, full-stack apps, and the Linux infrastructure they run on. Sixteen years of building software that stays up.

PythonGoPHPTypeScriptJavaC++PostgresLinux
monica@lfdln: ~
01 · selected work

Platforms I built, run, and keep running.

Three production systems, designed end to end: architecture, databases, security boundaries, deployment and the tests that keep them honest.

16years shipping software, since 2010
3platforms in production
4primary languages
B.S.computer science, UCSD
02 · how I work

From the schema to the server.

I take a request all the way: requirements, data model, API, interface, deployment, and the years of maintenance after.

$ design
Relational models that protect data integrity, APIs with clear contracts, and architectures split into modules with one job each.
$ secure
Least privilege by default: separate service users, restricted helpers, rootless containers, nothing running as root that doesn't have to.
$ test
Unit tests and Playwright end-to-end suites, plus verification runs on every install.
$ operate
Linux, systemd, nginx, Podman and Quadlet; backups, monitoring and documented runbooks.
03 · earlier work

Built in 2012, still in use in 2025.

In Winter 2012 I was an undergraduate tutor for CSE 21, Mathematics for Algorithm and Systems Analysis, taught at UC San Diego by Professor Ronald Graham, the mathematician known for Graham's number: counting, recurrences, decision trees and graph theory. I also designed and built the course website. Its template was still running CSE 21 thirteen years later.

04 · foundations

Neural networks, from scratch.

A computer science degree from UC San Diego, including a course where the networks were written by hand before the frameworks did it for you.

$ degree
B.S. in Computer Science, UC San Diego, 2016. Coursework in data structures and algorithms, discrete mathematics, data analysis and compiler design.
$ neural networks
CSE 190 with Professor Gary Cottrell, 2015. I implemented networks from scratch in Python, including a recurrent neural network and its training. The one exception was a convolutional network, built with Caffe.

Let's build something that lasts.

Open to engineering roles and independent projects.

✉ emailrésumégithub ↗photography ↗