Show HN: Pebble Falcon–autonomous right-sizer and carbon-aware K8s scheduler
What it Pebble Falcon:
Two long-running agents you deploy inside your k8s cluster.
> PerfectFit:
- Streams pod / VM / GPU utilization
- Calculates head-room
- Generates JSON patches to right-size CPU, memory and storage
- Optional human-approval step
> EcoAgent
- Polls real-time grid-carbon feed and re-configures GPU nodes and pods to maximize clean energy use.
- Re-queues batch jobs to the region with the lowest gCO₂/kWh that still meets your latency SLA
A lightweight web UI shows cost saved, watts avoided and a 30-day projection.
No data leaves the cluster; the only inbound feed is public carbon intensity.
Why we built it:
We kept seeing GPU nodes idling below 20 % while ETL jobs ran in high-carbon regions. Manual “rightsizing sprints” never caught up.
Pilot result (40-node GPU cluster, 30 days) 70% idle compute removed
4160 kg CO₂ avoided (≈ six NYC→London flights)
Pebble Falcon needs access to clusters so no public sandbox yet. Instead we recorded a 3-min overview of the agents.
Video Demo: https://youtu.be/DTpHxAmVrAo
However if you’re interested in trying it out, comment and we’ll get in contact and can add you to our pilot test: https://www.gopebble.com/sign-up
AMA about energy, cost and compute aware scheduling for AI workloads.