Inform
Close the month
Compare billed Azure cost for [cost group] in [month A] with [month B]. Rank the three largest service and subscription drivers, call out anomalies and unallocated spend, and tell me what to investigate next.
Ask Finn prompt library
Ask a scoped cost question. Get a reviewable investigation.
Start with a prompt built for FinOps, finance, engineering, or product work. Replace the bracketed details, paste it into Ask Finn, and keep the same scope when you verify the answer.
How to use the library
Replace the brackets before you ask.
Ask Finn does not inherit the top-bar filters. Put the period, Azure scope or cost group, and cost basis in the sentence. Add the decision you need at the end.
Use billed cost when you need the invoice view.
Use amortized cost when you want commitments spread across the period they cover.
Use Explorer when the question grows into a deeper saved analysis.
Prompts for
FinOps practitioner
Close the month, triage anomalies, fix allocation gaps, and keep the savings queue tied to evidence and ownership.
Inform
Triage an anomaly
For [cost group], compare billed Azure cost on [anomaly date] with the prior 30-day baseline. Identify the services and resources behind the increase, separate usage from rate effects, and list the evidence an owner should check.
Operate
Find allocation gaps
Show unallocated billed Azure cost for [month] by subscription and service. Rank the largest gaps, include the percentage of total spend, and identify missing cost-group or owner context when it is available.
Optimize
Rank open savings
Rank the open Azure savings recommendations for [cost group] by annual savings. Include monthly savings, the evidence, caveats, and owner when available, then suggest the first three reviews to schedule.
Prompts for
Finance and leadership
Question the budget, explain forecast movement, reconcile showback, and prepare a concise cost brief with a clear source trail.
Inform
Estimate a budget crossing
Using billed Azure cost for [cost group] from [month start] through [as-of date], when will spend cross the [budget amount] budget at the current run rate? Show actual spend, the forecast, and the assumptions behind the date.
Operate
Explain forecast movement
Compare the current amortized Azure forecast for [fiscal period] with the forecast published on [prior date] for [cost group]. Rank the changes by service, separate new run rate from one-off spend, and state which assumptions moved.
Operate
Review showback
Summarize billed Azure cost for [month] across [cost groups] and reconcile it to the total. Show allocated and unallocated amounts, shared-cost treatment, and any group whose month-over-month change needs an owner explanation.
Operate
Draft a cost brief
Create a five-point Azure cost brief for [month] using billed cost across [scope]. Include spend versus budget, the largest variance, material anomalies, open annual savings, and the one decision leadership needs to make next.
Prompts for
Engineering
Connect a cost change to a workload, test whether a deployment changed the run rate, and review recommendation evidence before action.
Inform
Check deployment impact
Compare daily billed Azure cost for [workload or cost group] during the seven days before and after [deployment date]. Rank the resources and meters that changed, and tell me whether the pattern looks persistent or temporary.
Inform
Trace a service driver
Why did billed Azure cost for [service] in [cost group] change between [period A] and [period B]? Break the variance down by subscription, resource, and meter, then list the first evidence the service owner should verify.
Optimize
Review right-sizing evidence
For the highest-value right-sizing recommendation in [cost group], summarize the current resource size, measured utilization window, proposed size, projected annual savings, and every caveat the engineer should check before changing it.
Operate
Prepare an owner handoff
Summarize the [anomaly or savings] investigation for the owner of [cost group]. Include the period, billed Azure impact, likely driver, supporting evidence, uncertainty, and a read-only checklist for the next review.
Prompts for
Product and project teams
Connect cost to a product or project outcome, watch AI and Copilot run rate, and distinguish a temporary spike from a new baseline.
Inform
Explain unit cost
Compare amortized Azure cost per [business unit] for [product cost group] in [period A] and [period B]. Show the change in total cost, usage volume, and unit cost, then identify which driver explains the difference.
Inform
Track AI and Copilot
Summarize billed AI and Copilot spend for [product cost group] from [start date] through [as-of date]. Rank model, workspace, and license drivers, forecast the end-of-month total, and flag the earliest budget crossing risk.
Operate
Review a project forecast
Using amortized Azure cost for [project cost group], compare actual spend through [as-of date] with the [project period] budget. Show the current run rate, forecast at completion, and the assumptions the project manager should revisit.
Optimize
Challenge the baseline
For [product cost group], compare billed Azure cost in [recent period] with the prior [baseline period]. Separate one-off anomalies from sustained service growth and explain whether the budget baseline should stay, rise, or be investigated further.
Before a number leaves CloudMonitor
Use Finn to find the answer. Use the reports to cite it.
The question and verification view should use the same period, cost group, and cost basis.
Ask one factual question
Finn is strongest when the prompt names the period, scope, measure, and decision.
Confirm report-bound numbers
Open The Ledger with the same scope before a value enters a close pack, forecast, or approval.
Continue deeper analysis
Move to Explorer when you need to shape, save, and share a multi-step analysis.
Keep the human decision
Finn can recommend a next step. Your owner and approval process decide whether any Azure change happens.
Start an investigation
Try the month-close prompt in the recorded demo.
The public sandbox opens directly on Ask Finn. Its four investigations are recorded and read-only, so you can see the evidence and follow-up pattern without a customer workspace.
Ask Finn Grounded in CloudMonitor data Read-only
What changed in July versus June, and what should I investigate next?
Which open savings should we prioritize?
Add the period, cost group, and billed or amortized cost basis.
See how Ask Finn carries a question into the evidence.
Open the recorded investigations directly, or review the agent’s product scope and trust boundary first.
