Today we’re excited to announce a new partnership between Pebble and WattTime to help organizations optimize cloud and Kubernetes workloads for lower energy use and lower carbon, without sacrificing performance.
Pebble is building an AI-powered, infrastructure-agnostic optimization platform that reduces compute waste, energy consumption, and carbon emissions. Our PerfectFit Agent right-sizes CPU/GPU/memory across clusters, and our EcoAgent safely participates in utility demand response by automatically adjusting compute loads during DR events.
WattTime is an environmental tech nonprofit best known for its marginal emissions data and Automated Emissions Reduction (AER) approach, signals that let software and devices shift flexible demand to the cleanest grid moments. Their API provides real-time, forecast, and historical grid emissions signals (including MOER: Marginal Operating Emissions Rate) so you can time energy use when the grid is cleanest.
Why this matters
Not all kilowatt-hours are created equal. Running the same job at different times can have very different carbon impacts depending on the marginal generator on the grid. That’s precisely what marginal emissions capture, and why they’re the right signal for load shifting and operational decisions.
What we’re launching together
Pebble + WattTime brings carbon intelligence straight into day-to-day compute operations:
Carbon-aware right-sizing & scheduling
Pebble consumes WattTime’s real-time and forecast MOER signals to guide when and where to scale clusters, run batch jobs, or schedule ML training, so you cut cost and emissions at once.AER-powered flexibility
With AER concepts, Pebble’s EcoAgent can automatically shift flexible workloads to cleaner five-minute windows, aligning with utilities’ DR events and minimizing emissions during peaks.Global coverage from day one
WattTime’s data signals cover grid regions worldwide, with hourly marginal emissions for nearly all global electricity consumption, ideal for organizations operating in multiple regions.
How it works (under the hood)
Region detection & enrollment
Pebble maps your clusters and cloud regions to the correct grid region via WattTime endpoints, then authenticates to pull emissions signals.Signals we use
We query real-time and forecast MOER (with five-minute granularity) plus historical data for baselining and reporting. These signals arrive via WattTime’s API and are designed for operational decisioning.Optimization loop
Pebble’s agents blend performance SLOs, cost, and marginal emissions to decide when to scale, queue, or shift jobs, then generate clear reports showing avoided emissions from timing and right-sizing.
Results you can expect
Lower compute spend via right-sizing and smarter scheduling
Lower kWh by eliminating over-provisioning
Lower tCO₂e by aligning usage to cleaner grid intervals (the exact metric WattTime’s MOER is built to inform).
For data teams & sustainability leaders
Whether you run on-prem Kubernetes, multi-cloud AI/ML, or high-throughput batch, you’ll get actionable insights and automated controls, plus credible accounting backed by the same marginal-emissions approach recommended for real-world impact.
Get started
Want to see carbon-aware compute in action?
Request a pilot with Pebble (PerfectFit + EcoAgent).
Together with WattTime, we’re making it effortless to choose cleaner, cheaper compute, and to prove it with auditable data.
We accelerate climate action by empowering businesses to reduce their carbon footprint. Our focus on transparency, accountability, and impact drives progress in carbon offsetting, renewable energy, ocean conservation, and biodiversity protection. Together, we build a sustainable future.