RTCF Prompt Builder - Conversation, Not Delegation

Prompt Builder

RTCF Framework

Advanced Techniques

Build Your Prompt

R - Role Who should the AI be? What expertise should it have?

T - Task What should the AI do? Be specific about the action.

C - Context What background information does the AI need?

F - Format How should the output be structured?

Generated Prompt

Fill in the RTCF fields and click "Generate" to build your prompt...

Tips

• You don't need all 4 elements every time --- use what's relevant
• RTCF is your conversation starter, not your final answer
• Follow up with refinements: "Make it more specific" or "Add relevant domain context"
• Remember to VET (Verify, Explain, Test) the output

The RTCF Framework

Role

"You are an experienced higher education analyst"

Task

"Identify the three strongest themes and two areas of concern"

Context

"Second-year unit, 85 responses, two campuses"

Format

"Present as a structured table with statistical summary"

Before and After

Weak Prompt:
"Help me write a marketing strategy."

RTCF Prompt:
"You are an experienced marketing strategist specialising in B2B SaaS (R). Develop a go-to-market strategy for a new project management tool (T). We have a small team, limited budget, and need to differentiate from Jira and Asana (C). Present as an executive summary followed by a phased 90-day action plan (F)."

RTCF in Practice

• You don't need all 4 elements every time --- use what's relevant
• RTCF is your conversation starter, not your final answer
• Follow up with refinements: "Make it more specific" or "Add industry context"
• Think of AI as a capable colleague, not a magic tool
• The best prompts invite conversation, not just commands

Advanced Conversation Techniques

Reverse Prompting

Let AI interview you to build context before giving recommendations. Surfaces assumptions you didn't know you had.

I'm working on a project that I need help thinking through.

Before giving me any advice, interview me first. Ask me questions one at a time to understand my specific context, constraints, and goals.

Start with your first question.

Multi-Perspective Analysis

Get expert viewpoints from different disciplines on your research question. Catches blind spots.

I'm investigating [your research question].

Act as my advisory panel and analyse this from multiple perspectives:

1. As a domain expert: Technical feasibility?
2. As a statistician: Best experimental design?
3. As an industry partner: Commercial viability?
4. As an ethics reviewer: Any concerns?

For each, give me 3-4 key points.

Devil's Advocate

Force AI to argue against your position. Overcomes confirmation bias and reveals hidden assumptions.

I believe that [your hypothesis].

Play devil's advocate. Give me the strongest possible arguments against this. Be harsh but fair. I want to know:

1. Why might this fail?
2. What am I probably not thinking about?
3. What alternative explanations exist?
4. What would a sceptical reviewer say?

Don't hold back.

Conversational Data Analysis

Upload data and explore it through conversation rather than a single prompt.

I've uploaded a [type] dataset from a [description] study.

What patterns do you see? Give me an initial overview of the data.

[Then follow up with:]
- Why might [observation] occur?
- Are there any unexpected trends?
- Create a visualisation of [specific aspect]
- What statistical tests would be appropriate?

VET Framework Check

Use after any AI interaction to push back and verify outputs.

You just told me [AI's claim].

Help me VET this:
- VERIFY: What sources support this claim? Can I find this independently?
- EXPLAIN: Can you break this down so I can explain it in my own words?
- TEST: What edge cases or conditions might make this wrong?

Literature Synthesis (NotebookLM)

After uploading papers to NotebookLM, use these prompts for synthesis.

Based on the uploaded papers:

1. What are the main findings across these papers?
2. Where do the authors disagree?
3. What gaps exist in this literature?
4. What methodological approaches are most common?
5. Suggest 3 research questions that emerge from these gaps.