FinOps Estimator
An interactive terminal utility for modeling CI/CD pipeline costs and Spot instance optimizations.
Who this Lab is for
Designed for
- Platform Engineers
- DevOps Engineers
- FinOps Practitioners
Use it when
- Evaluating cost reduction strategies
- Planning CI/CD infrastructure changes
- Estimating cloud compute budgets
A complete run, step by step
Input baseline metrics
Provide current build volume, failure rates, and compute hours.
Define cost structure
Input on-demand and spot instance pricing for your region.
Configure interrupt thresholds
Set your risk tolerance for build interruptions.
Calculate savings
View the modeled outcomes of various optimization strategies.
What you will need
Prepare the following information before starting. Use measured evidence where possible; defaults are examples and should not be treated as recommendations.
What the result tells you
Your report includes
- Accurate cost estimates based on pipeline volume
- Spot vs On-Demand savings models
- Waste reduction insights
How it is determined
Calculates annualized compute costs using exact AWS pricing formulas, factoring in retry overhead and spot interrupt probability.
Models compute costs directly using standard AWS EC2 on-demand and spot pricing structures, providing a strict baseline for comparing optimization scenarios.
Model assumptions
- • Spot instance interrupt rates align with historical AWS averages.
- • CI/CD build duration scales linearly with CPU/Memory.
- • Instance provisioning times are negligible compared to build times.
Authoritative references
Spot instance migration
Situation
Moving stateless build agents to spot instances with a 5% interrupt rate.
Result
Estimates a 60% reduction in compute spend while factoring in the 5% time waste from interrupted build retries.
Use the result with engineering judgement
- Does not factor in data transfer costs.
- Assumes spot capacity is always available.
Questions before you begin
How accurate is the spot pricing?
Spot pricing fluctuates, so the tool uses standard discounts (e.g. 70%) as a baseline model rather than real-time spot market pricing.
Does this work for Azure or GCP?
The pricing methodology applies to Preemptible VMs and Spot VMs on other clouds, though the discount thresholds differ.
What happens if spot capacity runs out?
You should maintain a fallback on-demand node pool to prevent total CI/CD outages during capacity crunches.
Ready to run FinOps Estimator?
Log in to open the full interactive workspace. Your completed result can be saved, revisited and exported as a private report.
Log in and start