VynOps
v0.1.0  ·  Open Source  ·  MIT License

VynCost — AI-Powered Cloud Cost Optimizer

Identify waste, rightsize workloads, and cut your cloud bill with AI-driven recommendations across AWS, GCP, Azure, and Kubernetes — in a platform your team owns.

Built on Next.js 15 · Groq Llama-4 · AWS Cost Explorer · GCP Billing · Azure Cost Management

Get Started Free View SourcePricing
AI anomaly detection
Rightsizing recs
RI & savings plans
Idle resource cleanup
Cost allocation
Auto-optimisation
Spend forecasting
Multi-cloud
Kubernetes FinOps
Budget alerts

Features

Cut your cloud bill intelligently

AI Cost Anomaly Detection

The AI monitors your cloud spend in real time and surfaces anomalies, sudden spikes, and trends — before your monthly bill arrives.

  • Spend spike detection: alerts when any service cost increases >20% in 24h
  • Baseline modelling: learns your normal spend pattern per service and region
  • Root cause attribution: which resource or change caused the cost jump?
  • Natural language cost summaries: "You spent $4,200 more this week — here is why"
  • Forecasting: projected end-of-month bill based on current trajectory

Rightsizing Recommendations

Identify over-provisioned compute, memory, and storage across AWS, GCP, Azure, and Kubernetes with concrete, safe downsizing recommendations.

  • EC2/GCE/Azure VM rightsizing based on actual 30-day utilisation patterns
  • Kubernetes pod resource request vs. actual usage gap analysis
  • RDS / Cloud SQL instance class rightsizing recommendations
  • Savings estimates in USD shown before any change is applied
  • One-click Terraform or kubectl patch commands to apply recommendations

Reserved Instance & Savings Plans

Analyse your on-demand spend and identify exactly which RIs or Savings Plans to purchase to maximise savings with minimum risk.

  • RI coverage analysis: what percentage of your compute is covered today?
  • Savings Plan purchase recommendations with break-even timeline
  • Convertible vs. standard RI comparison for your workload patterns
  • Multi-account RI sharing analysis (AWS Organizations)
  • Annual savings projection shown per recommendation

Idle & Orphaned Resource Detection

Find and clean up resources that are running but doing nothing — the silent budget killers.

  • Idle EC2/VM instances: running but 0 CPU for 7+ days
  • Unattached EBS volumes, disks, and snapshots older than 30 days
  • Unused Elastic IPs, NAT Gateways, and load balancers
  • Oversized Kubernetes nodes with <20% utilisation
  • Automated cleanup workflows with approval gate before deletion

Cost Allocation & Showback

Attribute cloud costs to teams, services, environments, and cost centres with tag-based allocation and showback reporting.

  • Tag-based cost allocation by team, service, environment, and project
  • Untagged resource detection and auto-tagging workflow
  • Showback reports: monthly cost per team or business unit
  • Chargeback integration: export cost data to finance systems
  • Budget alerts: notify team leads when budget thresholds are reached

Automated Optimisation Workflows

Schedule and automate cost-saving actions with approval gates, dry-run previews, and full audit trails.

  • Scheduled auto-scaling policies: scale down dev environments overnight
  • Spot/Preemptible instance migration for fault-tolerant workloads
  • Approval workflow: recommendation shown with risk rating before applying
  • Dry-run mode: see projected savings before committing any change
  • Full audit log: who approved what, when, and what savings resulted

Installation

Quick Start

From zero to first cost recommendations in under 15 minutes.

01

Clone and configure

git clone https://github.com/vynops/VynCost cd vyncost cp .env.local.example .env.local # Required: # VYNCOST_SECRET=<openssl rand -base64 32> # GROQ_API_KEY=<from console.groq.com> # At least one cloud provider: # AWS_ACCESS_KEY_ID + AWS_SECRET_ACCESS_KEY # GOOGLE_APPLICATION_CREDENTIALS=<path to SA JSON> # AZURE_CLIENT_ID + AZURE_CLIENT_SECRET + AZURE_TENANT_ID
02

Install and start

npm install npm run dev # Dashboard: http://localhost:3020
03

Connect your first cloud account

curl -X POST http://localhost:3020/api/accounts \ -H 'Content-Type: application/json' \ -d '{ "provider": "aws", "name": "production", "region": "eu-west-1" }' # Response: {"id":"acct_1","provider":"aws","status":"syncing"}
04

Connect your Kubernetes clusters

curl -X POST http://localhost:3020/api/clusters \ -H 'Content-Type: application/json' \ -d '{ "name": "production-eks", "apiUrl": "https://your-eks-api-server", "costRates": { "cpuPerHour": 0.048, "memGiBPerHour": 0.006 } }'

Environment Variables

Full reference for .env.local.

VariableDescriptionRequired
VYNCOST_SECRETRandom 32-byte secret for session signingrequired
GROQ_API_KEYGroq API key for AI cost analysis — free at console.groq.comrequired
AWS_ACCESS_KEY_IDAWS access key ID (Cost Explorer read access required)optional
AWS_SECRET_ACCESS_KEYAWS secret access keyoptional
AWS_REGIONSComma-separated AWS regions to scan (default: all)optional
GOOGLE_APPLICATION_CREDENTIALSPath to GCP service account JSON with billing read accessoptional
AZURE_CLIENT_IDAzure App Registration client ID (billing reader role)optional
AZURE_CLIENT_SECRETAzure client secretoptional
AZURE_TENANT_IDAzure tenant (directory) IDoptional
SLACK_WEBHOOK_URLSlack webhook for cost spike notificationsoptional
BUDGET_ALERT_THRESHOLDPercentage over budget to trigger alert (default: 80)optional
PORTHTTP port (default: 3020)optional

Supported Cloud Providers

Amazon Web Services
EC2, EKS, RDS, S3, Lambda, Cost Explorer API
Google Cloud Platform
GCE, GKE, Cloud SQL, GCS, Billing Export API
Microsoft Azure
VMs, AKS, Azure SQL, Blob, Cost Management API
Kubernetes (any)
Custom cost rates, pod/namespace breakdown
Free · Open Source · Self-hosted

Stop paying for waste

Connect your cloud accounts and start finding savings in minutes. No vendor lock-in.

Get Started View PricingTalk to us