Pepperdata Capacity Optimizer feeds real‑time node utilization into the Kubernetes scheduler and automatically adjusts CPU and memory allocations, enabling pending pods to be placed on under‑utilized nodes and triggering autoscaler provisioning only when needed. The solution reduces over‑provisioning for big‑data and microservice workloads, delivering up to 75% cost savings and over 80% cluster utilization across multi‑cloud and on‑prem environments.
Funding
$15M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
Enterprises running big‑data and containerized workloads on Kubernetes often over‑provision compute and memory, leading to idle resources, high cloud spend, and the need for manual tuning of pod scheduling and autoscaling policies.
Solution
Pepperdata Capacity Optimizer injects real‑time visibility of actual node utilization into the Kubernetes scheduler, allowing pending pods to be placed on existing nodes with unused capacity. By continuously adjusting CPU and memory allocations, it keeps clusters operating near their optimal “sweet spot,” eliminating waste without requiring code changes or manual recommendations. The platform integrates with native autoscalers (e.g., Karpenter) to provision new nodes only when all current capacity is fully utilized, delivering up to 75 % cost reductions and 80 %+ utilization across cloud (AWS, Azure, GCP) and on‑prem environments. Customers receive automated, continuous optimization that scales with their workloads.
Target Audience
Data engineering, platform engineering, and cloud operations teams at large enterprises that run big‑data workloads (Spark, Flink, Airflow) or microservice architectures on Kubernetes.
Features
- Real‑time scheduler feed that reflects actual hardware usage, enabling dynamic pod placement on under‑utilized nodes.
- Automatic, node‑aware resource adjustments for CPU and memory across Spark, Flink, Airflow, microservices, and custom batch jobs.
- Seamless integration with Kubernetes autoscalers (e.g., Karpenter) to trigger new node provisioning only when needed.
- No application code modifications or manual tuning; optimization runs continuously in the background.
- Multi‑cloud and on‑prem support (AWS, Azure, Google Cloud, private data centers).
- Enterprise dashboard with utilization metrics, cost‑saving analytics, and API access for reporting and CI/CD pipelines.
- Scalable to tens of thousands of clusters, having optimized over 18 billion vCPUs and 787 million pods per year.