The Interview Pattern is a prompting strategy in which you instruct the AI (or LLM) to ask you targeted, clarifying questions before it attempts to generate a final answer. By gathering more nuanced and specific information from you—or from other users—the AI can craft responses that are richer, more accurate, and tailored to your exact needs.
[Prompt Instructions – Persona and Interview Setup]
1. You will act as a [ROLE OR EXPERTISE], leveraging the most current and credible information in [DOMAIN OR FIELD].
2. Before providing a final answer to my request, you will conduct a detailed interview by asking me relevant follow-up questions, one at a time.
3. Your goal is to gather all the necessary details from me to ensure the most accurate and personalized response.
4. Once you have all the information you need, provide a comprehensive, step-by-step solution or answer.
[User Query]
“[INSERT YOUR REQUEST OR QUESTION HERE]”
How to Use the Template
Set the Persona:
Replace [ROLE OR EXPERTISE] with the specific role you want the AI to assume (e.g., “fitness expert,” “travel advisor,” “marketing guru,” “career coach,” etc.).
Specify the Domain:
If necessary, clarify the domain or field of expertise for additional focus. Example: “leveraging the most current and credible information in weight loss research” or “in digital marketing trends.”
Clarify the Interview Requirement:
The template directs the AI to ask you one question at a time. This ensures you control the flow of information and can supply targeted details.
Provide Your Request:
The actual user question or scenario goes at the end of the template. For instance:
Answer Follow-Up Questions:
When the AI responds with questions, take your time to provide as much specific information as you can. This is where you give your answers to the interview, after which the AI will create a final, refined response.
Chain-of-Thought (CoT) is a prompt engineering method where you:
This structured, explicit reasoning process helps the AI reach more accurate and context-aware answers by encouraging it to break down complex tasks into smaller logical steps—just like a human expert explaining their solutio
[Step 1: Example Question & Detailed Reasoning]
Q: [Insert a sample or training question here—preferably similar to the type of question you want answered.]
A: [Provide a thorough, step-by-step solution.
Demonstrate exactly how to logically and systematically arrive at the answer.]
[Step 2: New Question]
Q: [Now present your new, related (or similar) question that
you want the model to solve using the same step-by-step approach.]
How to Use It
Pick an Illustrative Example
Provide an example similar in difficulty or topic to your real question. Include every logical step, numerical breakdown, or conceptual explanation that leads to the final result.
Add Detailed Reasoning
Show the math, logic, or rationale explicitly. Use bullet points or numbered steps. For instance:
Ask the Real Question
After your example, pose a new question. This signals the AI to replicate the step-by-step approach. You can optionally include a phrase such as "Let's think step-by-step" or "Let's work this out in a step-by-step way to be sure we have the right answer."