EC2 & EBS waste alerts that live in Slack

Your idle EC2 instances are burning $140/month each

SaneCost detects idle instances and orphaned EBS volumes, then messages the owner directly in Slack. No dashboard to check. No login to remember. Just an alert where your team already works, with a button to act.

14-day free trial / 5-minute setup / Read-only AWS access
# aws-alerts
SaneCost
SaneCost 2:34 PM

Idle Instance Detected

i-0a4f2e8b1c in us-east-1 has been below 2% CPU for 6 hours.

Type m5.xlarge Cost $140/mo
SaneCost
SaneCost 2:34 PM

Owner @alice notified via DM.

SaneCost
SaneCost 2:35 PM

Orphaned Volume Detected

vol-0f3a7b2e9d in eu-west-1 has been unattached for 14 days.

Type gp3 Size 500 GB Cost $40/mo

Example Slack notifications from SaneCost

You're not alone

If your AWS bill feels wrong, it probably is. The industry wastes $44.5 billion a year on idle cloud infrastructure. Most teams don't even know it's happening.

Don't take our word for it:

What it does

Idle instances and orphaned volumes, flagged to the right person

01

EC2 & EBS Detection

Checks CPU utilization across your EC2 fleet every 30 minutes and flags unattached EBS volumes wasting storage costs. Filters out batch jobs, periodic workloads, and freshly-launched instances. You set the CPU threshold and observation window.

EC2 idle detection / EBS orphan detection / Owner extraction from tags / False positive filtering / Configurable thresholds
$ sanecost analyze i-0a4f2e8b1c
Instance i-0a4f2e8b1c · m5.xlarge · us-east-1a
CPU avg 1.2% over 6h · Network 0.3 MB/h · Threshold 5% (configured)
Owner alice@company.com (from tags)
▶ Verdict: idle — $140/mo recoverable

Your rules, not ours

You define what "idle" means for your infrastructure. Set your own CPU thresholds, observation windows, and instance exclusions. Every team's workload is different — SaneCost adapts to yours, not the other way around.

02

Slack Alerts

Posts to your team channel with full context — instance ID, type, region, daily cost, and how long it's been idle. If the instance has an Owner tag with an email, the owner gets a direct DM too.

Channel + DM delivery / Owner auto-detection / One-click stop

aws-alerts

Idle instance i-0a4f2e8b1c — $140/mo

+ owner gets a DM

DM to @alice

You own an idle instance — tap to stop it

03

Multi-Account Coverage

Scans EC2 instances and EBS volumes across all regions in every connected AWS account. Connect as many accounts as you need to one Slack workspace. Alerts are deduplicated — you'll never get spammed about the same resource.

Multi-account / Multi-region / Alert deduplication
$ sanecost status
Accounts 2 connected · Regions us-east-1, eu-west-1, ap-southeast-1
Last scan 2 min ago · Next in 28 min
Instances 147 monitored · Alerts 3 idle (deduplicated)
✓ Account healthy — no duplicate alerts sent

Setup

Three steps.
Five minutes.

From install to first alert in under five minutes.

Start now
01

Install the Slack app

One click to add SaneCost to your workspace. Choose which channel receives alerts.

02

Connect your AWS account

Deploy a CloudFormation template that creates a read-only cross-account role. Takes under 2 minutes.

03

Get your first alert

SaneCost scans immediately. Idle instances show up in Slack with full context and a stop button.

Security

Built for teams that care about access control

Read-only access

We only read CloudWatch metrics and EC2 metadata. Stop actions require your explicit click.

TLS encrypted

All data in transit is encrypted. API endpoints are secured and rate-limited.

No data stored

We don't persist your AWS data. Analysis happens in real-time, results go to Slack, nothing is retained.

IAM role-based

Standard AWS cross-account IAM role with least-privilege permissions. You control access.

Pricing

One plan. Flat rate.

14-day free trial, then a flat monthly fee. No per-seat pricing, no usage tiers, no surprises.

£19 /month per workspace

What's included

EC2 idle instance detection across all regions
Unattached EBS volume detection
Multiple AWS accounts per workspace
Owner DM notifications via tags
One-click actions from Slack
Configurable thresholds and snooze
Unlimited team members
Email support

How it works under the hood

Algorithms first, AI when it helps

Core detection is algorithmic — CPU thresholds, time windows, cost calculations. No black box. AI steps in only for edge cases: refining confidence on ambiguous patterns and extracting owners from messy tags. Every alert ships with raw data so you can see exactly why it was flagged.

You decide. We never act without your click. But we'll nudge you towards savings like nobody else can.

AI-assisted recommendations may occasionally be inaccurate. Every alert includes the data you need to decide — one glance, one click.

Your idle instances and orphaned volumes won't clean themselves up

Five minutes to install. First alert on the first scan. 14-day trial, no credit card.

Add to Slack 14-day trial · Read-only access