Presenters

Source

TamaGROTchi: Bringing Observability to Life (and Your Desk!) 👾✨

Remember those addictive digital pets from your childhood, the Tamagotchis? You know, the ones that had a mysterious knack for kicking the digital bucket, leaving you scratching your head wondering why? Well, prepare for a blast from the past with a modern, observability-infused twist! Chris and Nick from Grafana are here to introduce you to TamaGROTchi, the 2026 version that’s not just cute, but a powerful demonstration of how we can bring observability to the smallest of devices.

From Hackathon Fun to Real-World Impact 🚀

TamaGROTchi wasn’t born out of a grand product roadmap; it was the winning entry from Grafana’s quarterly hackathons! While many hackathon projects contribute to the main product (think AI Observability or the Grafana Assistant), Chris and Nick decided to create something a little more… cuddly.

“We decided to make a cute little pet,” Chris explains. This adorable digital dinosaur mascot, Grot, can now be taken anywhere, complete with a cell phone that alerts you via Slack when he’s hungry.

The Tech Behind the Tamagotchi 🛠️

At its core, TamaGROTchi is built on an ESP32 board. Every interaction, whether you’re feeding Grot or playing with him, is transmitted as logs, traces, and metrics via OpenTelemetry (Otel).

“It’s all fun and games, but actually a lot of our customers, they’re not actually using Kubernetes,” Nick points out. This is where TamaGROTchi’s real-world relevance shines. Many customers are building observability into small, hardware-constrained devices like robots in warehouses or POS terminals. These environments often have limited memory, unreliable internet, and can’t afford to run heavy collectors like the Otel Collector or Alloy, or even a full SDK.

This challenge inspired the team to experiment with sending raw Otel data. They’re creating JS on packets, sending them over HTTP, and directing them straight to Grafana Cloud. This approach demonstrates how to achieve observability even in the most resource-limited scenarios.

Explaining Observability to Your Mom (and Everyone Else!) 👨‍👩‍👧‍👦

“Our moms don’t know what we do at Grafana,” Nick humorously admits. As an Observability Architect and a Data Engineer, they spend their days immersed in complex concepts. TamaGROTchi, however, became the perfect, relatable tool to explain their work to their families.

Choosing the Right Telemetry Signals 📡

A key decision during development was which telemetry signals to use. Initially, they considered sending everything as a single log or metric. They even had a “sick metric” that was a simple zero or one, allowing them to easily count sick TamaGROTchis.

However, this limited their ability to answer crucial questions like: How did they get sick? Who let them get sick?

To gain deeper insights, they implemented:

  • Logs: Every button press is now a log event.
  • Tracing: This allows for a more comprehensive understanding of the interactions.

“We actually started implementing things like logs,” Chris explains. “Every button press that is a log. We’ve got tracing as well, and that let us build out a much better suite of dashboards.”

Visualizing Grot’s World: Dashboards and Knowledge Graphs 📊

Observing TamaGROTchi’s data involves a sophisticated setup:

  • Knowledge Graph/Entity Graph: This provides a high-level, top-down view of the TamaGROTchi ecosystem. Instead of Kubernetes pods and nodes, you see your TamaGROTchis! They’ve created custom entities for their use case, including SLOs (Service Level Objectives) that track when a TamaGROTchi is hungry or ill.
  • SRE Conventions: Tried-and-tested Site Reliability Engineering practices are in place. This includes SLOs like Mean Time To Feed, measuring how often a TamaGROTchi requests food and the response time. They also track RED metrics (Rate, Errors, Duration) for comprehensive performance monitoring.
  • Custom Dashboards: Beyond the high-level overview, custom dashboards offer granular views of individual metrics and logs. You can even dive into a single pet’s dashboard for a detailed look.

The AI Touch: Making Grot Smarter (and Grumpier!) 🤖

No GrafanaCON talk would be complete without a dash of AI! The team has integrated the Grafana Assistant with custom prompts. This AI now understands what a TamaGROTchi is, knows the importance of keeping it alive, and even gets “pretty annoyed when you neglect him.”

“It really gives a new name to postmortem,” Chris jokes, highlighting how the AI can interpret neglect and its consequences.

Come See TamaGROTchi in Action! 👋

Chris and Nick invite everyone to visit their booth at the science fair. You can:

  • See how the dashboards work – a fantastic learning tool!
  • Learn how to explain complex observability concepts to your family.
  • Play with the TamaGROTchis and experience the fun firsthand.
  • Enter the NFC raffle at the IoT booth for a chance to take a TamaGROTchi home!

TamaGROTchi is more than just a novelty; it’s a testament to Grafana’s commitment to making observability accessible, understandable, and even fun, from the largest enterprise systems to the smallest, most constrained devices. So, come on down, say hello, and maybe even save a digital pet’s life!

Appendix