Meeting A: Project Status Meeting
A single meeting note, good for practicing task framing and context preparation.
AI 101L Practice Center
Five judgment challenges: task framing, context selection, output review, Human-AI-Human workflow design, and Use Case readiness.
Info
AI 101L practice scenarios reuse AI 101's teaching-only simulated files. All names, companies, and figures are illustrative.
Warning
Use only the simulated files provided by the course. Unless authorized, never paste an organization's confidential data, personal data, or sensitive information into an AI tool; check your unit's rules before using real internal work data.
Related scenario files
Meeting B: Cross-Department Follow-up
A cross-department meeting note, good for practicing Human-AI-Human workflow design.
Customer Feedback: Raw Excerpts
Raw customer feedback excerpts, good for practicing output review and sensitive-data judgment.
Sales Report: Sales Data
CSV sales data, good for practicing Use Case readiness evaluation.
Practice challenges
P1: Task Framing Challenge
Your manager says "can you sort out Meeting A." Use the Trigger / Input / Task / Decision / Output / Reviewer framework to rewrite it into a clear task you could hand directly to AI.
Files used: Meeting A: Project Status Meeting
Steps
- Read all of Meeting A and judge which concrete tasks the notes can actually support.
- Write one sentence for each field: Trigger, Input, Task, Decision, Output, Reviewer.
- Combine the six fields into one work instruction you could give directly to AI.
- Write down three differences from the original vague instruction.
What you produce
- A one-sentence task frame for each of the six fields
- One complete work instruction
- Three differences from the vague instruction
P2: Context Selection Challenge
Meeting B contains several kinds of information: confirmed decisions, an unverified rumor, and another department's internal contact details. Classify each piece of information as needed, optional, sensitive, or not-allowed.
Files used: Meeting B: Cross-Department Follow-up
Steps
- List at least six distinct pieces of information from Meeting B.
- Tag each one as needed / optional / sensitive / not-allowed.
- For every sensitive or not-allowed item, write down why.
- Write the final context summary you would actually give to AI (only needed items and cleared optional items).
What you produce
- Classification of at least six information items
- Reasons for each sensitive/not-allowed item
- A final context summary
P3: Output Review Challenge
Suppose AI produced an analysis summary from Customer Feedback that includes a specific refund amount and a satisfaction score: numbers the source text never actually provides. Compare against the source and find everything worth questioning.
Files used: Customer Feedback: Raw Excerpts
Steps
- Read the Customer Feedback source carefully and note the concrete numbers and facts it actually contains.
- List the specific items in the AI output you believe have no traceable source (e.g. a number, a date, an owner).
- Tag each one as: unsupported claim, missing item, hallucinated date/owner, or uncertain content.
- Write how each one should be corrected.
What you produce
- A list of the facts actually in the source
- At least four questionable items with their tags
- A correction for each one
P4: Human-AI Workflow Builder
Using Meeting B (the cross-department follow-up) as the scenario, design a workflow from "meeting notes" to "the formal tracker being updated," tagging each step's executor and verification point.
Files used: Meeting B: Cross-Department Follow-up
Steps
- List at least five steps from the input to the final output.
- Tag each step's executor: Human, AI, or Tool.
- Write the Input, Action, and Output for each step.
- Add Verification to at least two steps, and describe when Escalation is needed.
What you produce
- A workflow table with at least five steps
- Executor tags for each step
- At least two verification points and their escalation conditions
P5: Use Case Readiness Challenge
Using the sales-tracking work shown in the Sales Report as context, judge whether it is currently ready to become a formal AI Use Case. This is a formative exercise with no single correct answer: what matters is your reasoning.
Files used: Sales Report: Sales Data
Steps
- Describe how often this work happens today and how much time it takes.
- List the current pain points and possible risks.
- Classify the work as: ready-for-experiment, needs-clarification, high-risk-review, or not-suitable.
- Write three reasons that support your classification.
What you produce
- A description of frequency and time spent
- A list of pain points and risks
- A readiness classification with three reasons
Info
This page does not reveal a correct answer or make an automatic business-value judgment; bring your reasoning to class or a later department interview for discussion.
The done mark is stored in this browser only (and synced once you join a cohort). Keep your own work saved separately.