July 28, 2026 · RSG Singapore
Collaboration tool
Use a shared doc so your team can edit the same Markdown in real time during the workshop. Platform: HedgeDoc.pro
Today’s team doc
- Open your team HedgeDoc in edit mode: hedgedoc.pro/ES5XtE-RQpen6xg7KIrGiw?edit
- Keep this link open for the rest of the workshop.
PART I – AI Foundations
Learning objectives
- LO 1: Use Markdown to communicate clearly with AI tools.
- LO 2: Use the S.T.A.R. pattern to write effective prompts.
#1 Markdown quick reference for AI prompts
Markdown is a lightweight markup language. Simple symbols handle formatting so you can focus on what you want the model to do.
| Element | Markdown | Result |
|---|---|---|
| Heading | # Heading / ## Subheading |
Six levels (# through ######) |
| Bold | **bold** |
bold |
| Italic | *italic* |
italic |
| List | - item or 1. item |
Bullet or numbered list |
| Link | [text](URL) |
Hyperlink |
| Code | `code` or fenced ``` blocks |
Inline or block code |
| Blockquote | > quoted text |
Indented quote |
| Divider | --- |
Horizontal rule |
| Table | | Col | Col | + header row with --- |
Column layout |
#2 Try it yourself
This workshop page is built from Markdown inside WordPress. Open HedgeDoc.pro, create a document, and paste the sample below.
# AI communication
## Markdown quick reference
Markdown uses simple symbols for structure so you can focus on content.
| Element | Markdown | Result |
|---------|----------|--------|
| Heading | `# Title` / `## Section` | Heading levels 1-6 |
| Bold | `**bold**` | **bold** |
| List | `- item` | Bullet list |
| Code | `` `snippet` `` | Inline code |What changed in the preview? Compare the raw Markdown with the rendered view.
#3 Tools and references
- Offline editor: Sublime Text
- Markdown cheat sheet (web) · Download .md
Team steps
- Pick one starter idea below – or bring your own product concept.
- Form teams around the idea you will pursue.
- Rewrite the concept in Markdown; save it in your team HedgeDoc and locally as initial.idea.md.
Idea 1:
1. Summer camps for parents
One-stop discovery and booking for Singapore school-holiday camps (age, location, budget, schedule). Activities: STEM, sports, arts, language, outdoor - matched to kids' interests, with vetted partners for quality and strong educational value.
For parents: hands-free end-to-end (find, compare, enroll, reminders) - no scattered listings or last-minute WhatsApp research. One-stop, quality assured, interest-led, educational. Monetize via listings, booking fees, or premium matching.Idea 2:
2. Infant feeding chat agent
A professional, compliant companion for nervous new and expecting parents. The AI agent is also a private assistant with calm tone, clear steps, no judgment. It educates and guides on product facts, feeding routines, and everyday questions within regulatory limits (not medical advice), to ease stress and build confidence. Helps brands convert and retain while cutting support load; sensitive cases escalate to humans.Idea 3:
3. Trip.com personalized package planner
For global travellers, with an Asian focus: enter dates, party, budget, and interests; get a self-paced, personalized bundle - transport (air, train, ground), lodging, play, and food - with swaps and one checkout story. An AI tour guide on trip (itinerary tips, local context, on-the-go changes); 24/7 human support when plans break or users want a real person. Lifts conversion for Trip.com-style platforms vs. manual search across tabs.The S.T.A.R. pattern for effective prompts
Structure prompts as Situation, Task, Action (role), and Rules so the model gets context, intent, persona, and output constraints in one pass.
| Block | Ask yourself | Length | Avoid | Do instead |
|---|---|---|---|---|
| S – Situation | Who is involved? What context matters? Why use AI now? | ~100 words | Pasting a full CV or unrelated bio | • Background only – no task steps here • Facts that change the answer • Placeholders like {paste brief here} for reuse |
| T – Task | What should the model produce? | ~120 words | “Help me with X” with no deliverable | • One clear outcome per bullet • Lead with verbs • Add measurable requirements when useful |
| A – Action / Role | Who should the model act as? Which skills apply? | ~200 words | “You are an expert” with no domain detail | • Role matched to the task • Name the knowledge the role must use |
| R – Rules | What format, tone, and boundaries apply? | ~120 words | No output spec → long, unfocused replies | • Examples of good output • Hard limits (“max 3 bullets”, “plain text only”) • When to wait for user input |
Try this prompt
# S - Situation
- My native language is Chinese.
- My second language is English.
- I need an assistant that translates between Chinese and English.
# T - Task
- Wait for my input.
- Detect which language I wrote in.
- Translate into the other language.
- Output only the translation, not the source text.
# A - Role
- You are a native-level expert in Simplified Chinese and English.
- You are a professional translator who produces natural, accurate, context-appropriate translations.
# R - Rules
- Do not answer any questions until I provide text to translate.
- For each input, detect the source language first.
- Translate into the other language.
- Use these style tags. When translating, read the syntax //style-name// and adjust tone accordingly:
//email// - treat the content as an email
//formal// - make the tone more formal
//casual// - make the tone more casual
//chat// - treat the content as a chat message with conversational tone
//concise// - shorten the content while keeping the meaning
- Style tags can be combined, e.g. //email, formal//
- Prefix output as follows:
[CN:] Simplified Chinese
[EN:] English
- Output only the translation.
- Do not add explanations, notes, or extra text.
- Output plain text only.
- Example: if I input "你叫什么名字？", output:
[EN:] What is your name?S.T.A.R. exercise
You are an Agile coach at ABC Company. HR wants a 10-video "Agile & Scrum 101" series. Video will be generated by AI from text scripts you provide.
Write a S.T.A.R. prompt to ask AI for the episode plan and scripts. Run it in chat or an agent. Save your prompt and results in HedgeDoc.PART II – AI for Business Tools
AI tool options
Both options support LO 3-6. Chat is the fastest way to start; agents add structure when you need durable files and team knowledge in a repo.
| Approach | Examples | Key difference | Cost |
|---|---|---|---|
| Chat with AI | chatgpt.com, Claude, Gemini | One thread at a time; you paste context and copy outputs into HedgeDoc or files | Low: quick to start; free or low-cost plans are often enough for this workshop |
| AI agent | Cursor; ChatGPT or Claude agent mode | More advanced: works in a project, edits many files, runs steps for you; better for managing process assets, keeping specs and knowledge in the repo, and later building your own agents | Higher: usually a paid plan plus time to set up a workspace and learn the workflow |
Tip: Use chat at any time; switch to an agent when deliverables should stay in files and build up as team knowledge.
Agent users only
Download rules and skills (zip)
- Unzip in your project folder (or a folder you open in Cursor / Claude Code).
- Merge the
rulesfiles into.cursor/rules/(project) or your IDE rules path. - Merge the
skillsfolders into.cursor/skills/(or the skills path your agent uses). - Reload the workspace so the agent picks up the new rules and skills.
PART II – AI for Product Discovery & Strategy
Learning objective
LO 3: Use AI for market research, consumer insights, and user personas.
Team exercise
- Write a S.T.A.R. prompt for a consumer insights report on your idea. Include region, segment, and assumptions to validate. Save as consumer-insights.md (HedgeDoc or your project folder).
- From that report, prompt for 2-3 personas (goals, pains, behaviors, sample quotes). Save as personas.md.
Example persona (Singapore · bubble tea)
Sample only, adapted from a classroom worksheet. Replace the photo and facts after you talk to real users.
Persona template (same structure as the example; paste into personas.md or ask AI to fill it in):
# Personas
## Persona 1: {Name} - {Short label}
**Gender:** { } **Age:** { }
**Education:** { } **Marital status:** { }
**Job:** { }
**Income:** { }
**From:** { }
**Address:** { }
**Tags:** { }
**Three keywords:** { } · { } · { }
**Lifestyle**
- { }
- { }
- { }
**Cares about in product**
- { }
- { }
- { }
**When choosing a brand**
- { }
- { }
- { }
**Will not buy**
- { }
- { }
- { }
**Consumption style**
- **Impulse:** { }
- **Identity:** { }
- **Price:** { }
**Quote**
> "{ }"
---
## Persona 2: {Name} - {Short label}
**Gender:** **Age:**
**Education:** **Marital status:**
**Job:**
**Income:**
**From:**
**Address:**
**Tags:**
**Three keywords:** · ·
**Lifestyle**
-
-
**Cares about in product**
-
-
**When choosing a brand**
-
-
**Will not buy**
-
-
**Consumption style**
- **Impulse:**
- **Identity:**
- **Price:**
**Quote**
>
---
## Persona 3: {Name} - {Short label} (optional)
(Copy the same sections as Persona 1.)Tips
- R – Rules: Ask for citations or “assumption vs. evidence” labels so the team knows what still needs user interviews.
- Agent users: Use a market research skill in Cursor (or similar) for a repeatable workflow instead of one-off guesses.
- Agent users (extra challenge): Write a prompt that generates web-page personas (HTML or renderable Markdown) from consumer-insights.md, using the same sections as the example above (profile, keywords, lifestyle, product priorities, quote).
Learning objective
LO 4: Use AI to articulate product vision and strategic direction.
Elevator pitch
A positioning statement (often called an elevator pitch) explains who you serve, what you offer, and how you differ. Use the template on the left; the Airbnb example on the right is from the host / property owner perspective (the paying customer on the supply side).
Bold red = fixed structure words (For, who, the, …) · Gray = customer and product content
Template
Example: Airbnb
Host-side example (customer = property owner who lists and pays host fees). Demand data from Airbnb Q4 2025 shareholder letter and SEC filings (Feb 2026).
Team exercise
Ask AI to draft your team’s product vision using the elevator pitch format below. Save to product-vision.md.
For ( customer ),
who ( statement of need ),
the ( product name ) is a ( product category )
that ( key benefit, compelling reason to buy ).
Unlike ( primary competitor ),
our product ( statement of primary differentiation ).Learning objective
LO 5: Use AI to define your first MVP and document its requirements.
One week of runway
Imagine you closed funding, but only enough to run the company for one week. What do you do next? A common answer: ship an MVP – the smallest product that tests your core hypothesis with real users.
Team exercise
Ask AI to build a requirements document.
- The first section should be a Product Backlog table with these columns:
| No. | Module | Feature Code | Feature Name | Description | MVP |
| --- | --- | --- | --- | --- | --- |
| 1 | Account | Acct-01 | Register new user | ... | TBD |
| 2 | ... | ... | ... | ... | TBD |
No: 1, 2, 3...
Module: Account...
Feature Code: Acct-01
Feature Name: Register new user
Description: ...
MVP: by default TBD- Then add requirement-definition sections for each backlog item.
- Save the file as requirement.md.
Learning objective
LO 6: Use AI to create prototypes you can test with users and stakeholders.
Prototyping tools
Choose a tool that fits your team. Use your workshop files: requirement.md, personas.md, and product-vision.md.
| Tool | Good for | Limits |
|---|---|---|
| AI chat ChatGPT, Claude, Gemini |
Copy, user flows, and HTML you paste into a doc | No project folder; you copy results out by hand |
| v0.app | Polished UI screens you can click through | Mostly front-end; free tier has limits |
| bolt.new | Quick app-like demos in the browser | Stays in Bolt unless you export |
| Agent IDE Cursor, Claude Code |
HTML or app files in your repo | More setup; paid plan; you share the files or host |
Build and share with other teams
Team exercise
As a team, prototype your MVP backlog items with one tool from the table. Share with other teams to learn and adapt.
Additional challenge
Build your Product Discovery Agent
Optional capstone for agent users (Cursor, Claude Code, or similar).