Sample reports

Read a whole report before you talk to us.

Three accounts, three shapes of waste. Open any one and you get the report a customer gets: every finding, what it costs per month, what to change, and how much work that change is — not a summary of one.

Synthetic These accounts are the analyzer's test fixtures, not customer data. The checks, pricing tables, and output are identical to a real run. Figures are estimates based on on-demand us-east-1 pricing, as labeled in each report.

AI-heavy startup

Inference on SageMaker and Bedrock, one account, nobody watching the GPU bill.

$6,373/mo identified

21 findings · 1 account

Remediation mix

13 config change 6 migration 2 rearchitecture
Open the report → PDF

Where the money is

AI & ML $4,874 FinOps baseline $1,499
$1,561Previous-generation GPU endpoint, p3 → g5
$1,028Real-time endpoint billing with zero invocations
$967Sparse-traffic endpoint that belongs on async inference

Data platform, 2-account org

A platform account and an ML account, swept together and reported as one.

$5,576/mo identified

13 findings · 2 accounts

Remediation mix

7 config change 5 migration 1 rearchitecture
Open the report → PDF

Where the money is

AI & ML $3,526 FinOps baseline $2,051
$1,050Bedrock on-demand spend past the commitment breakeven
$922Training jobs that fit managed Spot
$881Endpoint idle 86% of days

Lift-and-shift enterprise

A datacenter migration with no AI workloads at all — the value floor on its own.

$2,821/mo identified

18 findings · 1 account

Remediation mix

14 config change 3 migration 1 rearchitecture
Open the report → PDF

Where the money is

No AI/ML waste found FinOps baseline $2,821
$1,132On-demand EC2 with no Savings Plan coverage
$510NAT gateway data-processing charges
$320Cross-AZ data transfer between chatty services

Want this for your own account?

Deploy the read-only template, get the same report on your architecture. Nothing leaves your account, and the walkthrough is 30 minutes.

Book a walkthrough → How the analyzer works