The Transferable Skill

Everything you learned about prompting AI for code works for any domain:

Code Prompt:
"Refactor this function to use async/await, add error handling,
 and return a typed response object."

Documentation Prompt:
"Rewrite this technical doc to explain the concept to a non-technical
 product manager. Keep it under 500 words. Include a diagram description."

Data Prompt:
"Analyze this CSV of sales data. Find the top 5 products by revenue,
 identify month-over-month trends, and flag any anomalies."

The pattern is identical: Role + Context + Task + Constraints + Format.

AI for Documentation

Documentation is AI's sweet spot — it's tedious for humans but straightforward for AI:

Generate docs from code:

Read all the route handlers in /server/routes/admin/ and generate
an API reference document. For each endpoint, include:
- HTTP method and path
- Required headers (authentication)
- Request body schema with types
- Response schema with examples
- Error responses
- Rate limiting info if applicable

Keep docs in sync:

Compare the current API documentation in docs/api.md with the
actual route handlers in server/routes/. List any discrepancies:
- Endpoints in code but not in docs
- Documented parameters that don't exist in code
- Wrong response formats in docs

Write release notes:

Read the git log since tag v1.8.0. Generate release notes for v1.9.0.
Group changes by: Features, Bug Fixes, Breaking Changes.
Write for a technical audience. Include commit references.

AI for Data Analysis

You don't need to be a data scientist to analyze data with AI:

Quick analysis:

Here's a CSV with 10,000 rows of customer support tickets.
Columns: id, created_at, category, priority, resolution_time_hours,
         satisfaction_score, agent_name

Answer these questions:
1. Average resolution time by priority level
2. Which agent has the highest satisfaction scores?
3. Are resolution times trending up or down over the last 6 months?
4. What % of tickets are resolved within 4 hours by category?

Output the analysis as a markdown table + key insights.

Generate visualization code:

Write a Python script using matplotlib that creates:
1. Bar chart of tickets by category
2. Line chart of average resolution time per month
3. Scatter plot of resolution time vs satisfaction score
Save each chart as a PNG file. Use a clean, professional style.

AI for Automation

Automate repetitive tasks with AI-generated scripts:

File organization:

Write a bash script that:
1. Scans ~/Downloads for files older than 30 days
2. Moves PDFs to ~/Archive/Documents/
3. Moves images to ~/Archive/Images/
4. Moves everything else to ~/Archive/Other/
5. Logs what was moved to ~/Archive/move-log.txt
6. Add a dry-run flag (--dry-run) that shows what would happen
   without actually moving anything

Data transformation:

Write a Node.js script that reads all JSON files in ./data/,
combines them into a single CSV file, and:
- Flattens nested objects (user.name.first → user_name_first)
- Handles missing fields with empty strings
- Sorts by date field descending
- Outputs to ./output/combined.csv

Workflow automation:

Write a GitHub Actions workflow that:
1. Runs on every PR to main
2. Runs linting (ESLint) and tests (Jest)
3. Checks that test coverage is >= 80%
4. Posts a comment on the PR with the coverage report
5. Blocks merge if any check fails

AI for Communication

Draft professional communications in seconds:

Meeting summaries:

Here are my rough notes from today's architecture meeting:
- decided to go with postgres instead of mongo
- need to migrate by Q3
- john will lead the migration
- concerns about downtime during migration
- budget approved for $20k in AWS credits

Turn these into a professional meeting summary email.
Include: attendees section, decisions made, action items
with owners and deadlines, and open questions.

Technical proposals:

Write a one-page technical proposal for migrating our
authentication system from session-based to JWT.

Include: current state, proposed architecture, migration plan,
risks and mitigations, estimated effort, and success metrics.

Audience: VP of Engineering (technical but time-constrained).
Tone: confident, data-driven.

Building AI into Your Workflow

The key is making AI usage habitual, not occasional:

Daily tasks where AI saves 30+ minutes:

Task	Manual Time	With AI	Savings
Write tests for new function	30 min	5 min	83%
Draft status update email	15 min	2 min	87%
Analyze error logs	45 min	10 min	78%
Write API documentation	60 min	15 min	75%
Create bash automation script	40 min	5 min	88%
Summarize meeting notes	20 min	3 min	85%

The compound effect: 30 minutes saved daily = 2.5 hours/week = 130 hours/year = over 3 working weeks reclaimed.

The Prompt Engineering Mindset

What you've learned in this course applies everywhere:

1. Be specific — vague inputs produce vague outputs
2. Provide context — the AI can't read your mind
3. Set constraints — format, length, tone, audience
4. Iterate — first output is rarely final
5. Review critically — AI output is a draft, not gospel
6. Build templates — save time on repeated tasks
7. Teach the AI — AGENTS.md, system prompts, examples

AI is the most powerful tool you'll use in your career. The better you communicate with it, the more value you extract from it. And that skill only grows with practice.

---quiz question: What is the main transferable skill from AI-assisted coding? options:

• { text: "JavaScript programming", correct: false }
• { text: "Structured prompt engineering — Role + Context + Task + Constraints + Format", correct: true }
• { text: "Touch typing speed", correct: false } feedback: The prompt engineering pattern (Role, Context, Task, Constraints, Format) works identically for code, documentation, data analysis, automation, and communication. It's a universal skill for working with AI.

---quiz question: How much time can AI save on routine tasks like documentation and data analysis? options:

• { text: "About 10% on average", correct: false }
• { text: "75-88% on most routine knowledge work tasks", correct: true }
• { text: "No time savings — you still have to review everything", correct: false } feedback: AI typically reduces routine task time by 75-88%. While review is still necessary, the total time (AI generation + review) is dramatically less than doing the work manually from scratch.

---quiz question: Why should you make AI usage a daily habit rather than an occasional tool? options:

• { text: "To justify the subscription cost", correct: false }
• { text: "The compound effect — small daily savings add up to weeks of reclaimed time per year", correct: true }
• { text: "Because AI tools stop working if not used daily", correct: false } feedback: Saving 30 minutes daily compounds to over 130 hours (3+ working weeks) per year. The key is building AI into your daily workflow so the savings are consistent and automatic.