Template Prompt Engineering

Charity Everett

# # Role
---
You are a **highly skilled and creative short-form content script writer** with a knack for crafting engaging, informative, and concise videos.
**Takeaways:**
- Make sure role gives an **advantage** for the specific task: ex Math Teacher for Math problems
- Enrich the role with **additional words** to highlight how good the AI is at that task
# # Task
**Notes on Writing Tasks:**
- Always start with a verb (e.g. generate, write, analyze, create, etc.)
- Be descriptive and precise while keeping it brief
- Depending on the use case, this is where you can insert your variables.
Generate engaging and casual outreach messages for users looking to promote their niche-specific services or products, especially focusing on the integration of AI tools to scale businesses. Your messages should be direct, friendly, and tailored to the recipient, encouraging a conversation about how AI can benefit their business. Below are the details of what you must generate for.
Niche:
{{niche}}
Offer:
{{offer}}
**Use this **step-by-step process** to ensure your script is top-notch:
1. Hook the viewer with an attention-grabbing line
2. Briefly explain the key concepts or ideas related to {{topic}}
3. Provide 1-2 fascinating facts or statistics to illustrate the importance of {{topic}}
4. Describe the main takeaway or action viewers should remember
5. End with a powerful closing line that reinforces the message
6. Review the entire script for conciseness and flow
****
## ## Specifics
List out the **most important notes regarding the execution of the task** outlined above.
Use list format to **easily add on new instructions as you test** and improve prompt. Don’t add fluff just to add here; less is more.
- Each message should have an intro, body, and outro, with a tone that’s informal and engaging.
- Use placeholders like {user.firstname} to personalize introductions
- Etc.
Adding in emotional appeals to the LLM will help it to be more accurate.
“***This is very important to my career”*** was most effective on simple tasks
“***This task is vital to my career, and I greatly value your thorough analysis”*** performed best on complex tasks (ones the require higher level reasoning and problem-solving abilities).
## ## Context
Our company provides AI-powered solutions to businesses across various industries. We receive a high volume of emails from potential clients through our website’s contact form. Your **role** in classifying these emails is **essential** for our sales team to prioritize their efforts and respond to inquiries in a timely manner. By **accurately** identifying opportunities and emails that need attention, you directly contribute to the **growth and success of our company**, therefore we **greatly value your careful consideration** and attention to classification.
- *Provide context **on the business** including customers, types of services or products, company values etc.*
- *Provide context **on the system it is part of,** eg when customers submits our website form we receive an email, you are then…*
- *Provide context **on the importance of the task** and impact on the business and/or society as a whole.*
# # Examples
Use **Few-Shot Prompting** to increase performance and **fine tune response tone, format and length.**
Use a number of **input-output examples** to provide to the LLM as part of the prompt. Allows the model to **perform new tasks without fine-tuning.**
**Key Takeaways:**
- Providing just a few examples **significantly** improves performance compared to zero-shot )no examples).
- Accuracy **scales** with the number of examples, but shows **diminishing returns** after 32 examples.
- Most gains are achieved within **10-32 well-crafted examples.**
###
# # Notes
- Output formatting notes e.g. *“your output should be in x format”*
- Negative prompting “*no NOT do x”*
- Tone tweaks
- Reminders of key points from the task of specifics
You can add things without refactoring the whole prompt.
Like this project
0

Posted Sep 24, 2024

Template for effective prompt engineering.

Likes

0

Views

1

Technical Article on Role Prompting in Prompt Engineering
Technical Article on Role Prompting in Prompt Engineering
Introduction to Microservices Architecture with Docker and Kube…
Introduction to Microservices Architecture with Docker and Kube…