Instrumenting my espresso machine with OpenTelemetry

Man tracks every coffee shot like a crisis, and the internet is deeply divided

TLDR: A developer hooked his espresso machine up to software that tracks every detail of a coffee shot to troubleshoot bad brews. Readers were split between impressed nerds and people mocking it as ad-like, overengineered nonsense — which is exactly why everyone kept reading.

A programmer turned his home espresso machine into a full-blown data project, wiring it up so it records heat, pressure, flow, and timing for every shot like it’s a tiny factory under investigation. His big idea is simple enough for non-coffee obsessives: when a drink tastes bad, don’t guess — measure everything. So yes, this man effectively gave his coffee machine the same kind of monitoring tools companies use to watch websites and apps, all to figure out why one espresso came out sour.

But the real entertainment is in the comment section, where readers split into instantly recognizable camps. One side looked at the whole thing and screamed “blogspam” and “AI slop,” with multiple people openly accusing the post of reading like an ad in nerd cosplay. Another camp was less angry and more existential, with one commenter basically asking whether humanity has finally lost the plot: your coffee tastes bad? “Add some milk and get on with your life!” Meanwhile, the engineers wandered in and did what engineers do best — ignored the culture war and zoomed in on one niche implementation detail, with one person comparing the machine’s split-brain setup to high-frequency trading systems.

And then came the quiet killer of every flashy tech post: a reader noticed the code samples had no indentation. Brutal. So while the article wanted to be a love letter to measured, perfect espresso, the crowd turned it into a referendum on modern tech excess, ad-like writing, and whether we are all one touchscreen away from becoming completely ridiculous.

Key Points

  • The article presents espresso extraction as a measurable system with many interacting variables, including grind, pressure, temperature, humidity, and bean age.
  • A Gaggia espresso machine running the open-source GaggiMate ESP32 controller was instrumented to diagnose shot quality issues.
  • The observability setup uses OpenTelemetry for instrumentation, ClickHouse Cloud for storage, and ClickStack for ingest, querying, and visualization.
  • The author models fast-changing machine signals as metrics and each espresso shot as a trace with child spans for brew phases.
  • The implementation had to account for ESP32 hardware limits such as constrained RAM and the need to protect the real-time control loop.

Hottest takes

"Feels like AI slop writing" — Gigachad
"Add some milk and get on with your life!" — echelon_musk
"AI-written blogspam where the whole thing is an ad" — sen
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.