Build a Weekly Ops Review with Claude

Create an audited operations workbook and management brief with Claude.

Introduction

30 Second Summary

A polished weekly report can still send a manager toward the wrong decision. One unusual case can distort a headline number while the rest of the team's work stays hidden.

In this project, you will use Claude to turn synthetic service data into a verified weekly operations review pack. The workflow moves from a convincing average-only story to an audited workbook plus a concise management brief.

What You'll Build

Your finished pack gives a manager a verified workbook plus a concise PDF brief ready for a weekly decision meeting.

By the end of this project, you'll have:

A traceable workbook in Weekly_Ops_Review_v2.xlsx that lets you check every KPI against the source rows.
An outlier analysis that shows how one 40-hour case pulls North's average above its median.
A management brief in Weekly_Ops_Brief.pdf that links every numerical claim to audited evidence before recommending three prioritized actions.
Secret Mission: Add a labeled what-if scenario showing how resolving one extreme case changes North's metrics without altering the baseline.

Are there any prerequisites?

You need access to Claude on a Windows device. No spreadsheet software, extra account, or installation is required.

Before We Start

Before the analysis begins, define the management decision your review pack should support. This commitment keeps the project focused on verified evidence that prevents unsupported AI conclusions from driving service-operations priorities.

Verify Claude File Creation

The review pack only matters if a manager can receive it as a real file. Chat text alone cannot become the workbook or PDF brief you plan to hand over.

Claude can create downloadable files when its file capability is enabled. In this step, you'll prove that capability with a tiny workbook before relying on it for the full report.

In this step, get ready to:

Enable Claude's file-creation capability.
Create a test workbook with one text cell and one numeric cell.
Download the test workbook to your Windows device.

Check Claude's file-creation capability

The Code execution and file creation capability lets Claude generate downloadable spreadsheets. Checking it now removes a setup risk before you build the full review pack.

Return to Claude on your Windows device.
Open Settings in Claude.
Select Capabilities.
Find Code execution and file creation.
Check whether the setting is on.
Turn on Code execution and file creation if it is off.

You should now see Code execution and file creation switched on. That's the first dependency cleared: Claude is ready to create the workbook files this project needs.

Create the test workbook

A tiny workbook isolates file creation from the operations analysis. Its exact filename and cell values give you a concrete artifact to verify.

Start a new chat in Claude.

You'll see a blank conversation ready for the capability check.

Place the exact workbook request in the message box by using this prompt:

Create a tiny Excel spreadsheet named capability-check.xlsx. Put File creation works in cell A1 and the value 4 in cell B1. Give me the finished file to download.

What does this prompt test?

The filename capability-check.xlsx makes the generated artifact easy to identify.
Cell A1 holds the text File creation works.
Cell B1 holds the numeric value 4.
The request asks Claude to return a finished file that you can keep outside the chat.

Before you send the prompt, do you expect Claude to return chat text alone or add a downloadable workbook?

Send the message.

You'll see a downloadable file named capability-check.xlsx in the conversation. You've proved the important part: Claude can turn a precise request into a file you can keep.

Download capability-check.xlsx directly from the conversation.

The downloaded workbook contains File creation works in cell A1. It contains the numeric value 4 in cell B1.

Don't see a downloadable workbook?

Return to Settings in Claude.
Select Capabilities.
Confirm that Code execution and file creation remains on.
Send the workbook prompt again in the new Claude chat.

Help me troubleshoot why Claude did not return the requested downloadable workbook.

Your account has now produced a downloadable workbook. Next, you'll use this same chat to establish the operations baseline before any dashboard is built.

Validate the Operations Data

Your downloadable test workbook proves Claude can create a real file. Now you need to prove its analysis starts from the right records.

A manager needs a quick baseline before a polished report can be trusted. Missing rows or misunderstood blanks would distort every later metric.

You will make the synthetic operations rows the only source of truth. You will verify the record count plus the status totals. You will also confirm that open-case blanks are intentional.

In this step, get ready to:

Supply the synthetic operations dataset to the existing Claude chat.
Constrain Claude to validate only the supplied rows.
Reconcile Claude's response with the expected baseline.

Supply the operations source data

A CSV dataset stores each case as one row beneath a fixed set of columns. This synthetic dataset lets you practise operational analysis without sharing private business data.

Switch back to the Claude chat that contains your capability check.
Paste the operations source into the message composer by copying the block below:

Case_ID,Team,Priority,Status,Resolution_Hours,SLA_Hours,Reopened,Customer_Rating
N-101,North,High,Closed,4,8,No,5
N-102,North,High,Closed,6,8,No,5
N-103,North,High,Closed,8,8,Yes,4
N-104,North,High,Closed,40,8,No,2
N-105,North,Normal,Closed,10,16,No,4
N-106,North,Normal,Closed,12,16,No,4
N-107,North,High,Open,,8,No,
N-108,North,Normal,Open,,16,No,
S-201,South,High,Closed,7,8,No,5
S-202,South,High,Closed,8,8,No,4
S-203,South,Normal,Closed,9,16,No,4
S-204,South,Normal,Closed,10,16,Yes,3
S-205,South,Normal,Closed,11,16,No,4
S-206,South,Normal,Closed,12,16,No,4
S-207,South,Normal,Closed,13,16,No,4
S-208,South,Normal,Open,,16,No,

What does this dataset capture?

Each Case_ID identifies one service case.
The Team, Priority, and Status columns describe how each case is classified.
The Resolution_Hours and SLA_Hours columns support later speed and compliance checks.
The Reopened and Customer_Rating columns capture service quality signals.
The open rows deliberately leave Resolution_Hours and Customer_Rating blank.

Send the pasted dataset to Claude.
Confirm your sent message starts with the Case_ID header.
Confirm your sent message ends with the S-208 row.

Dataset looks truncated?

Compare the first line of your message with the eight-column header in the source block.
Paste the complete dataset again if the S-208 row is missing.

Help me check whether my pasted operations dataset contains every source row.

Ask Claude to validate the rows

A validation request defines the boundary for Claude's analysis. It also explains how to interpret the blank fields before any calculations begin.

Before you send the validation request, do you expect more closed cases or open cases? What should Claude do with blanks that belong to open cases?

Ask Claude to validate the supplied rows by sending this request:

Treat the CSV-formatted rows above as the only source of truth. Do not create a file yet. Confirm the column names, count the data rows, and report total, closed, and open cases. Treat blank resolution time and rating values on open cases as intentional. If information is missing for any other reason, say so rather than guessing.

How does this request protect the analysis?

The source-of-truth instruction prevents Claude from adding outside assumptions.
The file restriction keeps this checkpoint focused on validating the data.
The blank-value instruction preserves the difference between an open case and missing information.
The uncertainty instruction tells Claude to identify genuine gaps instead of guessing.

Reconcile the baseline

AI output verification means comparing a generated answer with a known fixture. Here, the fixture is the source dataset you supplied.

Before you compare the response, do you expect blank values on open cases to change the closed-case count?

Read Claude's response.
Confirm the column list includes Case_ID, Team, Priority, Status, Resolution_Hours, SLA_Hours, Reopened, and Customer_Rating.
Confirm the source includes N-101 through N-108.
Confirm the source includes S-201 through S-208.
Confirm Claude reports 16 total cases.
Confirm Claude reports 13 closed cases.
Confirm Claude reports 3 open cases.
Confirm Claude describes the blank resolution and rating values on open cases as intentional.

You should see 16 total cases. The status split should be 13 closed cases plus 3 open cases.

You should also see eight North cases plus eight South cases. Claude should preserve the intentional blanks in Resolution_Hours and Customer_Rating for open cases.

You have completed the essential quality-control checkpoint. Every later metric now starts from a reconciled source.

Counts don't match?

Compare Claude's reported row count with the 16 source rows in your pasted dataset.
Check that all three open rows are still labeled Open.
Send the validation request again after restoring any missing source rows.

Help me reconcile Claude's case counts with the operations source data.

Your operations dataset is reconciled. Next, you will turn these validated rows into a first-pass workbook.

Build the First-Pass Review

The validated source now gives you a reliable baseline of 16 cases. Your next goal is to turn those rows into a report that a manager can scan quickly.

A polished dashboard can make a single comparison feel conclusive. In this step, Claude deliberately uses average resolution time alone to create a convincing first pass before you audit its conclusion.

In this step, get ready to:

Generate a first-pass workbook from the validated operations data.
Download Weekly_Ops_Review_v1.xlsx from the existing Claude conversation.
Trace the closed rows behind each team's average.

Generate the average-only workbook

Average resolution time compresses several closed cases into one team-level KPI. This creates the quick comparison used in the first-pass dashboard.

Prepare Claude to create the first-pass workbook by copying this prompt:

Using only the supplied operations data, create an Excel workbook named Weekly_Ops_Review_v1.xlsx. Include a Raw Data sheet, a KPI Summary sheet, a Team Comparison sheet, working formulas, and one clear chart. For this first pass, compare team speed using average resolution hours for closed cases and state which team looks faster. Keep open-case resolution and rating cells blank. Return the finished workbook and summarize its main conclusion in chat.

What does this request control?

The source restriction keeps Claude grounded in the validated operations rows.
The Raw Data, KPI Summary, and Team Comparison sheets separate the evidence from the summary.
The closed-case average creates one simple measure of team speed.
The blank-cell instruction preserves the intentional gaps in open cases.

Paste the copied prompt into the message box in the existing Claude chat.
Press Enter to send the request.

You should see a downloadable Weekly_Ops_Review_v1.xlsx file in the conversation. Claude should also summarize which team looks faster by average.

Workbook missing or incomplete?

Check that file creation remains enabled if Claude returns only chat text. Reuse the exact request above if the workbook omits a required sheet or chart.

Help me diagnose the missing or incomplete workbook.

Download the first-pass workbook

Downloading the workbook preserves the first-pass result before the analysis changes. It also gives you a separate artifact to compare with the audited version later.

Download Weekly_Ops_Review_v1.xlsx from the generated file card in the existing Claude conversation.
Confirm that your browser shows Weekly_Ops_Review_v1.xlsx as a completed download.

Good progress. Your first-pass operations workbook is now saved as a separate portfolio artifact.

Trace the team averages

A team average becomes more useful when you can trace the records included in it. This inspection exposes the denominator behind each headline number.

Before you ask Claude, which team do you expect the single average comparison to favor?

Prepare the workbook inspection request by copying this prompt:

List the sheets you created, the average resolution time for each team, and the exact rows included in each average.

Why trace the included rows?

The row lists show whether Claude excluded open cases from the speed calculation. They also make each average reproducible from the source data.

Paste the copied inspection request into the message box in the existing Claude chat.
Press Enter to send the request.

You should see the Raw Data, KPI Summary, and Team Comparison sheets in Claude's response.

North should use 6 closed rows. Its average should be 13.33 hours.

South should use 7 closed rows. Its average should be 10.00 hours.

Claude should identify South as faster by average. This is the intended first-pass conclusion.

The result is preliminary. Its management story has not been audited against the broader operational evidence.

Do the averages differ?

Check that Claude used only rows where Status is Closed.
Check that blank Resolution_Hours values from open cases were excluded.
Ask Claude to recalculate from the validated source rows if either team count differs.

Help me reconcile the team averages.

That is the first pass complete. You have a polished workbook with a conclusion that still needs proof. Next, you will audit its metrics against the source rows.

Audit and Correct the Metrics

Your first-pass workbook gives you a polished team comparison. Its average-only conclusion identifies South as faster.

That conclusion hides North's lower typical result plus one extreme delay. In this step, you will use Claude to reconcile each KPI against the raw rows before creating a separate corrected workbook.

In this step, get ready to:

Audit every KPI against the source rows.
Explain the outlier effect through an average-versus-median comparison.
Create a separate corrected workbook with traceable metrics.

Audit the first workbook

An average uses every closed case. One outlier can pull it away from most results.

A median identifies the middle result after the resolution times are ordered. Comparing both measures reveals whether one extreme case is shaping the conclusion.

Before you ask for the audit, do you expect North's average and median to tell the same story?

Send the following request in the existing Claude chat:

Audit the first workbook against the raw rows. Define completion rate as closed cases divided by all cases. Define SLA compliance as closed cases where Resolution_Hours is less than or equal to SLA_Hours divided by all closed cases. Define reopen rate as closed cases marked Yes divided by all closed cases. Calculate average and median resolution hours using closed cases only. For every percentage, show the numerator and denominator. Print a reconciliation table in chat before creating a replacement file.

What does this audit request do?

The definitions lock each KPI to a specific population.
The numerator and denominator expose the calculation behind every percentage.
The closed-case rule keeps blank open-case resolution values out of the time metrics.
The reconciliation table creates a checkpoint before Claude generates another file.

Claude should print a reconciliation table in the chat before offering a replacement workbook. This table becomes your first visible evidence that the calculations were checked.

Missing definitions or denominators?

Resend the complete audit request without shortening it.
Check that the validated source rows remain available earlier in the same chat.
Ask Claude to stop before file creation if it skips directly to a workbook.

Use Help me get Claude to produce the complete reconciliation table. if the audit remains incomplete.

Reconcile the audit against the source rows

AI output verification means checking generated analysis against known source facts. Here, the reconciliation table becomes the evidence trail for each management claim.

Confirm the overall count is 16 total cases.
Confirm the closed count is 13 cases.
Confirm the open count is 3 cases.
Confirm completion is 13/16 (81.25%).
Confirm SLA compliance is 12/13 closed cases.
Confirm the reopen rate is 2/13 closed cases.
Confirm the closed-case average customer rating is 4.00.

Why do denominators matter?

A percentage without its denominator can hide which records were included. The fraction 12/13 proves that SLA compliance uses closed cases only.

The same audit trail applies to completion and reopen rates. A manager can trace each headline percentage back to the source rows.

Confirm North has 6 closed cases.
Confirm North has 2 open cases.
Confirm North shows 13.33 average hours / 9.00 median hours.
Confirm North's SLA compliance is 5/6.
Confirm South has 7 closed cases / 1 open case.
Confirm South shows 10.00 average hours / 10.00 median hours.
Confirm South's SLA compliance is 7/7.
Trace North's average-versus-median gap to N-104 in the source rows.

The 40-hour N-104 case raises North's average. North's 9.00-hour median shows the lower typical resolution time.

South's average matches its median at 10.00 hours. Its 7/7 SLA result also shows stronger compliance.

You have now exposed the exact weakness in the first workbook's story. The corrected interpretation is grounded in speed, consistency, and compliance.

Create the corrected workbook

A separate workbook preserves the original report as evidence of the first conclusion. The new version adds definitions, reconciliation checks, and a balanced team comparison.

Before you create the corrected workbook, which parts of the first conclusion do you expect the full evidence to change?

Send the following correction request in the existing Claude chat:

Create a corrected workbook named Weekly_Ops_Review_v2.xlsx without overwriting version 1. Include Raw Data, KPI Definitions, Reconciliation, Team Comparison, and Dashboard sheets. Use formulas where practical, label every denominator, highlight N-104 as the outlier, and explain that North has the lower median while South is more consistent and has better SLA compliance. Return the file and summarize every correction from version 1.

What does this correction request protect?

The new filename preserves Weekly_Ops_Review_v1.xlsx as the original analysis.
The KPI Definitions sheet documents the calculation rules.
The Reconciliation sheet preserves the audit trail.
The revised comparison separates typical speed from consistency.
The highlighted outlier shows why the first conclusion changed.

Download Weekly_Ops_Review_v2.xlsx from the conversation.
Keep Weekly_Ops_Review_v1.xlsx unchanged.
Review Claude's summary of the corrections.

No corrected workbook to download?

Confirm that Claude finished processing the file request.
Check that the response contains a downloadable Weekly_Ops_Review_v2.xlsx attachment.
Resend the complete correction request if Claude returns chat text without a file.

Use Help me troubleshoot the missing corrected workbook. if the attachment still does not appear.

Before you send the final check, do you expect the average and median to support the same team-speed conclusion?

Verify the corrected workbook by sending this request:

List the sheets in Weekly_Ops_Review_v2.xlsx. Confirm that formulas are used where practical. Confirm that every denominator is labeled. State how N-104 is highlighted. Then print the final audit table in chat using only the source rows and the values in the workbook. Show overall results. Show North results. Show South results. Include total cases, closed cases, open cases, completion rate, SLA compliance, reopen rate, average customer rating, average resolution hours, and median resolution hours where applicable. Show every percentage as a numerator and denominator. Finish with the corrected team comparison. Retract any value that cannot be traced to the source rows.

What should the final audit show?

The workbook lists Raw Data, KPI Definitions, Reconciliation, Team Comparison, and Dashboard sheets.
The overall row shows 16 total cases / 13 closed cases / 3 open cases.
Completion shows 13/16 (81.25%).
Overall SLA compliance shows 12/13.
The reopen rate shows 2/13.
The closed-case average customer rating shows 4.00.
North shows 6 closed cases / 2 open cases / 13.33 average hours / 9.00 median hours / 5/6 SLA compliance.
South shows 7 closed cases / 1 open case / 10.00 average hours / 10.00 median hours / 7/7 SLA compliance.
The workbook highlights N-104 as the 40-hour outlier.
The corrected narrative states that North has the lower median.
The corrected narrative states that South is more consistent.
The corrected narrative states that South has better SLA compliance.

Does a final value differ?

Compare the disputed value with the reconciliation table from earlier in the chat.
Check that resolution metrics use closed cases only.
Check that open-case resolution values remain blank.
Ask Claude to retract any result that it cannot trace to a source row.

Use Help me trace the mismatched KPI back to the source rows. if you need help locating the mismatch.

Your corrected workbook now tells a traceable management story without erasing the first version. Next, you will turn this verified evidence into a concise PDF brief.

Create the Management Brief

The audited workbook gives you reliable calculations. Those calculations become useful when a manager can grasp the operational risk quickly.

A concise brief connects that evidence to a decision. In this step, you will use Claude to create a PDF with traceable management actions.

In this step, get ready to:

Create a concise PDF management brief from the audited workbook.
Trace every numerical KPI claim to its supporting workbook sheet.
Check the boundary between evidence and interpretation.

Create the evidence-backed PDF

The brief uses the verified workbook as its evidence boundary. This keeps each recommendation connected to the operations data you audited.

Return to the existing Claude chat.
Create the management brief by sending this prompt:

Using only the verified values from Weekly_Ops_Review_v2.xlsx and the source rows in this chat, create a concise PDF named Weekly_Ops_Brief.pdf. Include an executive summary, five verified KPIs, the average-versus-median finding, one limitation, and exactly three prioritized actions. Recommend investigating N-104, triaging the two open North cases, and reviewing the two reopened cases. Do not add external benchmarks or invented causes.

How does this prompt control the brief?

The source boundary limits the brief to the audited workbook values plus the original rows.
The required sections give the manager a consistent reporting structure.
The three named actions connect the analysis to specific operational follow-up.
The final restriction blocks external benchmarks plus unsupported causes.

Monitor Claude until the file-creation response completes.
Download Weekly_Ops_Brief.pdf directly from the conversation.
Open Weekly_Ops_Brief.pdf from your browser's downloads list.
Confirm the PDF includes an executive summary.
Confirm it includes five verified KPIs.
Confirm it includes the average-versus-median finding.
Confirm it includes one limitation.
Confirm it includes exactly three prioritized actions: investigate N-104, triage the two open North cases, plus review the two reopened cases.

You should see a downloadable PDF containing each required section. The three prioritized actions should appear as a focused management plan.

No PDF or missing sections?

Return to Settings in Claude.
Select Capabilities.
Confirm Code execution and file creation remains on.
Resend the exact prompt above if the response contains chat text without a downloadable file.
Ask Claude to regenerate the PDF if any required section is missing.

Help me troubleshoot why Claude did not create the complete PDF brief.

You now have the management-facing deliverable. Its recommendations are grounded in the workbook you already reconciled.

Trace every numerical claim

A source trail makes a numerical claim auditable. It lets a manager challenge a number without rebuilding the analysis.

Before you send the next prompt, ask yourself whether every number in the PDF has a clear source.

Audit every numerical claim by sending this request:

For each numerical claim in the PDF, list the value, its numerator and denominator where applicable, and the workbook sheet that supports it. Retract any claim you cannot trace.

What does this audit request test?

Each percentage must expose its numerator.
Each percentage must expose its denominator.
Each numerical claim must name a supporting workbook sheet.
Any claim without a source must be withdrawn.

Review the claim trace in Claude's response.
Compare each listed value with Weekly_Ops_Review_v2.xlsx.
Confirm every applicable claim names its numerator.
Confirm every applicable claim names its denominator.
Confirm every numerical claim names a supporting workbook sheet.

You should see only values already present in the audited workbook. Every numerical claim should lead back to a named sheet.

Did Claude flag an unsupported claim?

Treat the unsupported claim as withdrawn.
Ask Claude to remove the claim from the PDF.
Download the corrected copy as Weekly_Ops_Brief.pdf.
Repeat the trace request against the corrected PDF.

Help me remove an unsupported claim from the management brief.

Check the evidence boundary

Numbers can be traceable while a narrative overstates what they prove. The workbook shows N-104 as a 40-hour outlier.

The cause of that delay remains unknown. Before your final check, ask yourself whether the brief claims to know that cause.

Find the N-104 discussion in Weekly_Ops_Brief.pdf.
Confirm it reports the 40-hour resolution time.
Confirm the cause remains unknown.
Confirm the brief compares average with median for team resolution time.
Confirm South is described as more consistent.
Confirm South's stronger SLA compliance appears as evidence.
Confirm the brief contains exactly the three prioritized actions from your original request.

What should the final brief prove?

North should show 13.33 average resolution hours.
North should show 9.00 median resolution hours.
South should show 10.00 average resolution hours.
South should show 10.00 median resolution hours.
North should show 5 of 6 closed cases meeting SLA.
South should show 7 of 7 closed cases meeting SLA.
The cause of N-104's 40-hour resolution should remain unknown.
The action list should investigate N-104, triage the two open North cases, plus review the two reopened cases.

That closes the reporting loop. Your management brief now connects every numerical claim to audited evidence.

Secret mission

Test an Outlier-Reduction Scenario

Your audited pack explains actual performance. Now test how resolving one extreme case changes the operational picture without altering the verified baseline.

Clean Up Your Resources

Choose whether to keep your downloaded files, set them aside for later, or delete them entirely. This Claude Free project creates no ongoing charges.

Resources you used:

The file-creation test is stored in capability-check.xlsx.
The review workbooks are Weekly_Ops_Review_v1.xlsx, Weekly_Ops_Review_v2.xlsx, and Weekly_Ops_Review_Scenario.xlsx.
The management brief is stored in Weekly_Ops_Brief.pdf.

Keep everything running

No action is required. Choose this if you want to keep reviewing the baseline, corrected analysis, management brief, or hypothetical scenario.

Keep all five downloaded files in their current folder.
Preserve the existing Claude chat so its source rows and reconciliation trail remain available.
Use Weekly_Ops_Review_v2.xlsx as the verified baseline when reviewing Weekly_Ops_Review_Scenario.xlsx.

Pause - I'll come back to this later

No process needs to be stopped. Choose this if you want to step away while keeping the complete analysis available.

Use your Windows file manager to confirm that all five downloaded files remain in their current folder.
Leave the existing Claude chat unchanged so its synthetic source rows remain available.
Close Claude when you finish reviewing the files.

Delete - I don't want to use this again

Remove every downloaded project file from your Windows device. Choose this if you no longer need the local review pack.

Use your Windows file manager to locate the folder where your browser saved the project downloads.
Select capability-check.xlsx, Weekly_Ops_Review_v1.xlsx, Weekly_Ops_Review_v2.xlsx, Weekly_Ops_Review_Scenario.xlsx, and Weekly_Ops_Brief.pdf.
Move the selected files to your Windows recycle area.
Empty the recycle area to remove the selected files permanently.
Search your Windows device for each of the five filenames.
Move any remaining copy found by the searches to your recycle area.
Empty your recycle area again after removing any extra copies.

You should find no remaining copies. That confirms every downloaded project artifact is gone from your Windows device.

Nice Work!

Your weekly operations review pack is complete. You transformed a 16-row service dataset into evidence for a manager's next decision.

What you learned:

Used Claude to build a downloadable operations workbook from structured service data. Made each KPI traceable to visible source rows. Supported each summary with working formulas.
Created an audited metric trail for every management KPI. Used average-versus-median analysis to expose the effect of N-104. Verified SLA compliance against the closed-case denominator. Replaced the preliminary conclusion with a balanced team comparison.
Produced a concise PDF management brief with five verified KPIs. Grounded every numerical claim in workbook evidence. The first action investigates N-104. The second action triages the two open North cases. The third action reviews the two reopened cases.
Secret Mission: Added a controlled what-if scenario that models N-104 at 8 hours. Kept every baseline value unchanged. Labeled every modeled result as hypothetical. Showed a hypothetical North average of 8.00 hours. Showed a hypothetical North median of 8.00 hours. Showed hypothetical North SLA compliance of 6/6.

Ready to quiz yourself?