Overslaan naar inhoud

Kema Labs

Case

Analyse the 2026 financial performance of Voltaris Testing Group using AI as a critical Finance & Control sparring partner.

Management wants to understand why financial performance differs from budget, which developments require attention, and what actions should be considered.

Assignment 1: Understand the Data
Explore the financial dataset with Copilot, identify data-quality issues and formulate relevant management questions.

Assignment 2: Build the Management Dashboard (Self-Ask)
Use AI to determine the right KPIs, generate Excel formulas and build a one-page dashboard comparing budget and actual performance.

Assignment 3: Challenge the Numbers (Analysis Tree)
Investigate the most important deviations, explore alternative explanations and identify which conclusions are supported by the data.

Assignment 4: Explore think-cell AI
Explore how think-cell AI can turn key financial findings into clear management visuals and where human judgement is still required.

Assignment 5: Draft the Management Report
Combine the dashboard, analyses and recommendations into a concise first draft of a management report for senior management.


Assignment 3: Understand the Data

Case: You are part of the Finance & Control team of Voltaris Testing Group. Management wants to understand the 2026 financial performance of its six laboratories compared with budget.

Goal: Use Copilot to understand the dataset, assess its quality and determine which management questions the data can answer.

Steps in the assignment:

  1. Download Voltaris_Financial_Performance_2026.xlsx
  2. Upload the Excel file to Copilot and ask it to explain the structure, variables and available business context.
  3. Ask Copilot to identify missing values, unusual observations and other potential data-quality issues. Check its findings yourself in Excel.
  4. Ask Copilot to propose 5 relevant Finance & Control questions that could be answered with this dataset, but do not analyse the financial performance yet.

Result: A clear understanding of the dataset, a checked list of data-quality issues and 5 management questions that will form the basis for the dashboard.

Assignment 4: Scenarioplanning opstellen

Case: Management wants a clear one-page overview of the 2026 performance of the six Voltaris laboratories.
Goal: Use Copilot to build the calculations and visualisations needed for a management dashboard.

Steps in the assignment:
  1. Use the questions from Assignment 1 and ask Copilot which KPIs are needed to answer them.
  2. Use self-ask to let Copilot break the dashboard task into smaller analysis questions.
  3. Ask Copilot to generate the required Excel formulas and add them to the file. Check the calculations yourself.
  4. Ask Copilot which charts best show the key differences between budget and actuals and build the dashboard in Excel. 

Result: A one-page Excel dashboard with relevant KPIs and visualisations for management.



Assignment 5: Challenge the Numbers

Case: The dashboard shows several significant deviations from budget. Management wants to understand what may be driving these results.
Goal: Use Copilot to explore possible explanations, challenge assumptions and identify what additional information is needed. 

Steps in the assignment:
  1. Select 2–3 notable findings from your dashboard.
  2. Ask Copilot to create an analysis tree with possible explanations for each finding.
  3. For each explanation, ask what evidence supports it, what could contradict it and what information is still missing.
  4. Discuss possible management actions with Copilot and decide which conclusions are sufficiently supported by the data. 

​Result: A structured set of insights, hypotheses and possible actions that can be used in the management report.


Assignment 6: Challenge the Numbers

Case: You already have a dashboard and several key financial insights. Now explore how think-cell AI can help turn these into clear management visuals. 
Goal: Discover the possibilities and limitations of think-cell AI when creating charts from financial data. 

Steps in the assignment:
  1. Select 2–3 relevant KPIs or findings from your analysis.
  2. Use think-cell AI to generate different chart options for the same insight
  3. Experiment with prompts for chart type, emphasis, titles and management message.
  4. Compare the outputs: what does the AI do well, where does it make weak choices and what still requires your judgement?
  5. Choose the strongest visual and add it to the management report. 

Result: A better understanding of think-cell AI, plus 1–2 management-ready visuals for the final report.


Assignment 7: Draft the Management Report

Case: Management now wants a concise written summary of the 2026 financial performance and the main areas that require attention.
Goal: Use Copilot Pages to turn your analysis into a clear and structured management report. 

Steps in the assignment:
  1. Upload or add the relevant outputs from the previous assignments as context.
  2. Ask Copilot Pages to create a draft with an executive summary, key variances, possible explanations, risks and recommended actions.
  3. Check whether every conclusion is supported by the data and adjust unsupported claims.
  4. Refine the tone and structure so the report is concise, professional and suitable for senior management. 

Result: A first draft of a management report that can be reviewed and further refined by the Finance & Control team.