From an AWS tool to a real multi-cloud FinOps platform

A month after launching in beta, the biggest change to PruneOps is this: it has stopped being an AWS tool and become a real multi-cloud dashboard.

What's there now

PruneOps now connects to and audits five different environments from the same dashboard:

• AWS — broad coverage: EC2, RDS, DynamoDB, ElastiCache, OpenSearch, S3, Lambda, VPC, KMS, Secrets Manager, and many more (33 resource types in total).
• OVH — Public Cloud instances, volumes, Object Storage, and DNS zones.
• DigitalOcean — Droplets, volumes, Spaces, and Load Balancers.
• Cloudflare — Zones, Workers, R2 Buckets, and Pages Projects.
• Kubernetes — workload metrics via a lightweight in-cluster agent, with no inbound access required.

Each provider has its own pricing module, its own detection logic, and its own recommendations — this isn't a thin generic layer bolted on top of every API, it's real per-service support.

Why this matters for FinOps

Most organizations don't live on AWS alone. There's usually some project on OVH for cost reasons, a service on Cloudflare for the edge, maybe a self-managed Kubernetes cluster. Waste doesn't stop at a provider's border — and partial visibility means partial decisions.

With this change, the same logic that already worked for AWS (continuous inventory, real metrics instead of estimates, recommendations with the numbers attached) now applies to every connected provider.

What's next

The natural next step is broadening coverage further: more services within the providers already connected, and exploring new ones (Google Cloud, Azure, Hetzner, Vultr, and Linode are the clearest candidates right now). Alerts and notifications, configurable budgets, and cross-provider cost allocation by tag are also on the radar.