Oopps.ai – Cost Optimization from Recommendations to Auto-Apply

Connect — See — Save

Oopps.ai works in three steps:

1. Connect in 5 minutes. Provide a Yandex Cloud service account — the platform automatically discovers all clusters, virtual machines, network infrastructure, and related resources — and starts collecting usage and state data.

2. Full infrastructure visibility. You see not just costs, but the state of everything: which resources are idle, where tags are missing for cost allocation, which clusters don’t meet basic security requirements. Real numbers instead of abstract charts — this deployment requests 4 CPU, uses 0.3.

3. Automatic optimization. The platform generates rightsizing recommendations based on actual consumption and applies them for you. Two options:

   • GitOps — Oopps.ai creates a Merge Request in your GitLab. You review, merge, and changes roll out through your CI/CD. Full control.
   • In-cluster agent — an agent inside the cluster applies changes directly. For teams that need speed without extra pipelines.

Why Not Do It Manually

Manual Kubernetes optimization means spreadsheets, kubectl top, guessing limits, and hoping nothing breaks. Common problems:

• Requests are 5–10x too high. Developers set them “just in case” — you pay for air.
• No consumption history. A point-in-time snapshot doesn’t show peak loads — recommendations based on it are risky.
• One-time optimization. You do a review, fix things — a month later it’s all back. You need a process, not a one-off.

Oopps.ai solves all three: analyzes consumption over time (not at a single point), generates safe recommendations accounting for peaks, and applies them automatically — continuously, not once.

Platform Capabilities

Cloud Resource Inventory and Governance

• Automatic inventory: VMs, disks, networks, public IPs, load balancers, K8s clusters
• Unused resource detection: detached disks, empty load balancers, orphaned IP addresses
• Governance scoring: missing tags, auto-generated names, unowned resources — every issue is visible and scored

Analytics and Cost Allocation

• Costs by cluster, namespace, workload — down to individual container level
• Cost distribution by team and project

Alerts and Reports

• Rules: sudden cost spikes, budget threshold, end-of-month overspend forecast, unused resource detection, cluster health score drop, pods without resource requests, overprovisioned resources, new optimization recommendations
• Regular digests (daily / weekly / monthly): costs by connector, unused resources, top savings recommendations
• Channels: Telegram, Email, Webhook — configured per organization

Cluster Security Audits

• Checks across 7 criteria: resource requests/limits, liveness/readiness probes, privileged containers, network policies, resource quotas
• Cluster health score with historical tracking
• Risk visualization by namespace — immediately see where attention is needed

Rightsizing and Automatic Optimization

• CPU and Memory recommendations based on real metrics over time
• Auto-apply via GitOps (GitLab MRs) or in-cluster agent
• Auto mode: the platform continuously optimizes resources without manual intervention

Multi-Cloud

• Yandex Cloud, Cloud.ru (SberCloud), MTS Cloud, Selectel, Azure
• All clusters and cloud resources in a single interface

1. Sign up at https://app.oopps.ai
2. Connect Yandex Cloud in your dashboard — instructions: https://oopps.ai/docs/connectors/yandex
3. Wait for data to load — first savings recommendations will appear within minutes.

For onboarding questions: support@oopps.ai

Cut Kubernetes costs without manual review
A team of 5 developers deploys 40+ services in Managed Kubernetes. Resource requests are set “just in case” — actual usage is several times lower. Oopps.ai analyzes metrics, generates rightsizing recommendations, and creates a Merge Request in GitLab with corrected requests/limits. The team reviews and merges — changes roll out through their existing CI/CD.

Automatic optimization without setting up pipelines
No GitOps, no time for setup — but the cluster needs optimization right now. The Oopps.ai in-cluster agent installs into the cluster and applies recommendations directly. Everything works out of the box — no pipelines, no MRs, no manual manifest edits.

Allocate cloud costs across teams
The CFO wants to understand how much each product team spends. Oopps.ai breaks down costs by namespace, workloads, and tags — down to individual containers — and maps them to teams. A weekly digest arrives by email.

Find forgotten resources you’re paying for right now
After a migration, the cloud is left with stopped VMs, detached disks, unused public IPs. Each one costs money but doesn’t show up in standard reports. Oopps.ai automatically finds these resources, estimates their cost, and notifies you via Telegram.

Cluster security audit in 60 seconds
Before a production release, you need to verify the cluster meets basic requirements: resource limits are set, no privileged containers, network policies are configured. Oopps.ai checks across 7 criteria and shows a health score with per-namespace detail.

Technical support is provided by the Oopps.ai team:
– Telegram: @ooppsai_support
– Email: support@oopps.ai
– Documentation: https://oopps.ai/docs