AI for Data Analysis Capstone Exercise
Deliberate Academy Editorial Team
Reviewed for accuracy and professional relevance
You're 10 lessons in — don't lose your progress.
Sign up free to save where you are and earn a verified certificate when you pass.
- Apply skills from across this course in a single realistic professional scenario
- Produce a concrete, role-relevant deliverable using AI tools
- Self-assess your output against professional quality criteria
This course has taken you from AI-augmented analytical workflows through natural language querying, data cleaning, visualization, storytelling, model evaluation, advanced statistical support, dashboard automation, and governance. The throughline across all of it is the same: AI accelerates the mechanical parts of data work, but the analyst's judgment determines whether the output is trustworthy. This capstone puts that principle under real pressure.
You are handed a messy dataset and a hard deadline. Your job is not just to produce a summary. It is to produce a defensible one: one where you can account for every step, flag every risk, and hand the output to an executive with confidence.
Capstone Exercise
End-of-Day Executive Summary: From Messy Sales Data to Board-Ready Findings
Context
You are a data analyst at a regional retail business. At 9am your manager sends you a spreadsheet of the last quarter's sales data across 12 store locations. The data has issues: some stores reported weekly instead of monthly, two locations have missing values in the revenue column for March, one store's figures appear to include returns already deducted while others do not, and a column labelled 'units sold' contains a mix of integers and text entries. Your manager needs an executive summary by 4pm covering the quarter's top performing locations, the key revenue trend, and one data quality caveat. You cannot fix the source data, only work with what you have.
Your Task
Write a five-step AI prompt sequence that guides an AI tool through the full analytical workflow for this dataset. Step one: data cleaning and normalisation instructions. Step two: calculation of summary statistics (total revenue, average by location, top three performers). Step three: trend identification across the quarter. Step four: drafting the executive summary paragraph. Step five: a disclosure statement for the two known data quality issues that will be included at the foot of the summary. For each of the five prompts, write the prompt text and then a risk annotation of one to two sentences identifying what you, as the analyst, must verify before trusting that output.
Your notes (optional)
Deliverable
A five-step prompt sequence with each prompt written out in full, plus a risk annotation on each step stating what the analyst must verify before using that output in the final summary.
Each of the five prompts in this capstone has to carry a risk annotation. What does the course say a risk annotation is actually for?
Select one answer.
- Describing data quality issues in your prompt is not optional: it is the step that determines whether the AI's cleaning instructions are safe to apply
- AI-generated summary statistics can be structurally correct but logically wrong: always trace the calculation back to how edge cases were handled
- A risk annotation is not a disclaimer, it is a professional accountability step that proves the analyst controlled the output rather than forwarded it
- Governance starts at the prompt level: how you frame the task determines whether the output is auditable later
Complete all lessons to take the free exam
Pass the exam to earn your AI for Data Analysis — Certified AI Practitioner — a verifiable certificate you can share on LinkedIn.