TutorialsAnalyze a Dataset with AI
Take a dataset, run the 9-step workflow, and produce a one-page management dashboard + action recommendations.
Following the 9-step workflow learned in class, find a dataset and run the full process from data to report. The completion standard is a one-page management dashboard + 3 action recommendations.
Course Learning Assistant
First check whether SkillHub is installed. If not, follow https://skillhub.cn/install/skillhub.md to install SkillHub (CLI only), then install the 通用数据分析流程 skill. If already installed, directly install the 通用数据分析流程 skill.
Course Materials
8 Prompt Templates
Covering data exploration through QA checklist (see Lesson 4: Self-Practice Templates).
5-Piece Prompt Formula
Role + Task + Background + Format + Constraint — the underlying formula for all templates.
6 Background Elements
Business, time, fields, definition, question, target — to stop AI from guessing.
Assignment: Analyze a Dataset with AI
Find a dataset → run the 9-step workflow → produce a one-page dashboard + action recommendations.
Completion criteria:
- Find a dataset (sales detail, expense records, survey, etc. — at least 50 rows)
- Use Template 1 (Data Exploration) for data profiling and quality check
- Use Template 2 (Sales Analysis) or the relevant scenario template to run the full analysis
- Use Template 7 (Action Recommendations) to generate a one-page dashboard + 3 improvement recommendations
- Use Template 8 (QA Checklist) to review all 9 items and confirm output quality
Step 1: Prepare Data
Find a dataset, spend 30 seconds verifying it: where it came from, time period, row count. Desensitize if it contains sensitive info.
Step 2: Data Profiling
Use Template 1 to have AI survey the data — report only, no analysis. Confirm null values, duplicates, and anomalies before continuing.
Step 3: Write the Prompt
Use the 5-piece formula + 6 background elements to write the prompt for metric calculation + chart generation.
Step 4: Build Report
Use Template 7 to generate a one-page management dashboard + action recommendations for management.
Step 5: QA Review
Use Template 8 to go through all 9 QA checklist items. Spot-check key numbers to confirm AI didn't fabricate.
Submission Requirements
Include Original Data
Submit the data file used for analysis (desensitized version), with source and row count noted.
Include Prompts
Attach your filled-in prompts (desensitize real info) and the tool name you used.
Include Analysis Report
Attach the AI-generated one-page dashboard + action recommendations, and your QA checklist results.
Write One Reflection
Which step took the most time? Where did AI save you the most? How would you improve next time?
No need to get it perfect on the first try — the key is to run the 9-step workflow end to end. Experience the rhythm of "fill in background → AI produces draft → you verify and approve." Feel free to share your work in the community group — the instructor will answer questions there.
Self-Practice Templates & Pitfall Guide
8 timeline-ordered prompt templates (data exploration → sales analysis → expense analysis → funnel → survey → chart styling → action recommendations → QA checklist), plus 3 general red lines + 4 data-analysis-specific red lines.
Making Talking-Head Short Videos with AI
6 high-frequency scenarios x 3 hands-on exercises x 5 prompt templates x 1 agent — one formula plus one agent to produce a shoot-ready talking-head script in 3 minutes.