About
Operations first, engineering second — in the useful order
I'm Rey Francis Talandrata, an automation engineer from the Philippines. I didn't come to automation from a computer-science lecture — I came to it from inside a business, watching hours disappear into work a machine should do.
The short story
For the past several years I've built and operated the automation layer for a US nationwide white-glove logistics company: document intelligence, call attribution, media pipelines, reporting, Slack bots, and the AWS infrastructure they run on. Not prototypes — production systems that people rely on daily, migrated across servers, monitored, and documented.
That origin shapes how I build. I don't start from "what's technically interesting" — I start from "what eats your team's hours, and what's the smallest system that gives those hours back." Then I write it in boring, dependable code and document it well enough that I could disappear and it wouldn't matter.
In 2026 I opened up my toolkit: thirteen open-source automation tools, each one tested, documented, and runnable in a minute. They exist so you can judge my work by reading it, not by believing this paragraph.
How I build
- Python & Node.js — plain, readable, dependency-light code
- Claude Code daily — I develop with AI agents natively, including building MCP servers that let Claude work with real business data
- APIs, webhooks & scraping — if it has an endpoint or an export, it connects
- AWS · Linux · systemd · cron — deployed on infrastructure you own, with monitoring and runbooks
- SQL & data pipelines — clean data first; automation on dirty data is just faster mistakes
- Where no-code fits — I'll recommend the simple tool when it's genuinely the right answer, and code when it isn't
Principles the code is built on
Boring is a feature
Every dependency is a future 3am problem. I reach for the standard library first, proven services second, and shiny things only with a reason. Systems built this way are still running years later.
Draft, don't fire
Automations that talk to your customers start in draft mode with human review. Trust is earned by a system, not assumed — a misfired email costs more than the minutes it saved.
Honest numbers only
Every metric on this site comes from a real system or a runnable demo. No invented percentages, no fake testimonials, no logos of companies I haven't worked with. If that costs me flashiness, good — you want the person who won't inflate your numbers either.
The practical details
| Location | Philippines — long experience working US hours with US teams |
| English | Fluent, written and spoken; async-first communicator |
| Engagements | Project-based builds or ongoing part-time; long-term preferred |
| Code | github.com/rtalandr18 |