Unveiling the Power of Prompt Chaining in AI

Prompt chaining is a revolutionary technique in AI that unlocks the true potential of large language models (LLMs) like Copilot, GPT-4, Jurassic-1 Jumbo, or Google Gemini. It allows you to break down complex tasks into smaller, more manageable steps, guiding the LLM through a series of prompts to achieve the desired outcome. Imagine it as handing a chef a recipe one step at a time, ensuring a delicious final dish.

Benefits of Prompt Chaining:

  • Simplified Instructions: One of the primary advantages of prompt chaining is that it allows us to write less complicated instructions. Instead of trying to express a complex task in a single prompt, we can break it down into smaller, more straightforward steps.
  • Tackling Complex Tasks: By dividing a large task into smaller, more focused prompts, LLMs can handle intricate problems they might struggle with in a single go.
    • Example: Imagine writing a legal document summarizing a complex contract. A single prompt might lead to a generic summary. With chaining, you could prompt for key clauses, risk factors, and then combine them into a final summary.
  • Focused Troubleshooting: Prompt chaining also enables us to isolate parts of a problem that the LLM might have difficulty with. If we encounter issues or inaccuracies in the responses, we can pinpoint the specific prompt in the chain that needs adjustment.
    • Example: A scenario of writing a legal document summarizing a complex contract using prompt chaining, as mentioned in the previous explanation. In this example, prompt chaining allows for focused troubleshooting:
      • Prompt 1: Summarize the key clauses of the contract.
      • Prompt 2: Identify any risk factors mentioned in the contract.
      • Prompt 3: Combine the summaries from Prompt 1 and 2 to create a comprehensive contract summary.
      • Imagine the final output (summary) doesn’t accurately reflect the risk factors. By analyzing the prompts and the corresponding outputs, we can isolate the issue:
        • If Prompt 1’s summary accurately captures the key clauses, and Prompt 3 correctly combines the summaries, then the problem likely lies in Prompt 2.
        • We can revisit Prompt 2, refine it to ensure it specifically targets risk factors, and re-run the prompt chain to see if the issue resolves.
  • Incremental Validation: Another benefit of prompt chaining is the ability to check the LLM’s output in stages, rather than waiting until the end. This incremental validation allows us to assess the correctness of responses as we progress through the prompts.
    • Example: Generating a social media post with an image.
      • Prompt 1: Find an image of a cat wearing a hat.
      • Validation: Here, you can visually validate the LLM’s response. Does the image indeed show a cat wearing a hat? If not, you can adjust the prompt (e.g., “Find a high-resolution image of a fluffy cat wearing a birthday hat”) and try again.
      • Prompt 2: Write a funny caption about the cat’s hat.
      • Validation: Read the caption generated by the LLM. Does it make sense in the context of the image (cat wearing a hat) and is it humorous? If not, you can provide more specific guidance (e.g., “Write a funny caption from the cat’s perspective about its stylish hat”) before proceeding.
  • Improved Accuracy and Control: Each prompt refines the LLM’s understanding, leading to more accurate and relevant outputs. You can steer the process in the desired direction with each step.
    • Example: Creating a product description. Start with prompting for the product’s features, then its benefits, and finally combine them into a compelling description.
  • Enhanced Creativity: Prompt chaining allows you to build upon the LLM’s initial creative spark.
    • Example: Generate a story idea. Start with a character and setting, then prompt for a conflict, and finally a resolution, building a more intricate plot.

Industries Harnessing Prompt Chaining:

  • Content Creation: Generate product descriptions, marketing copy, and even different sections of a blog post, all within a consistent style and tone.
  • Customer Service: Develop chatbots that can handle complex inquiries by chaining prompts based on user input, leading to a more natural and helpful conversation.
  • Software Development: Generate code snippets or even entire functions by providing step-by-step instructions through prompts.
  • Education: In educational settings, prompt chaining can be used to create detailed and step-by-step explanations of complex concepts.

Challenges and Solutions:

  • Prompt Design: Crafting effective prompts is crucial. Ambiguous prompts can lead to inaccurate results. Crafting effective prompts requires a deep understanding of the AI’s language model
    • Solution: Use clear and concise language. Provide specific examples when necessary to guide the LLM.
  • Maintaining Context: As the chain grows, ensuring all prompts are consistent with the overall goal can be tricky.
    • Solution: Break down the task into smaller, well-defined sub-tasks with clear transitions between prompts.
  • Avoiding Prompt Dependency: There’s a risk of creating a chain where later prompts are overly dependent on the success of earlier ones. 
    • Solution: Design the chain in a way that each prompt can stand on its own to some extent, reducing the dependency on previous prompts.
  • Evaluation: Measuring the success of prompt chaining can be subjective. The outputs from one prompt to the next can sometimes be inconsistent.
    • Solution: Establish clear evaluation criteria based on the desired outcome. This could involve human evaluation or comparing outputs to pre-defined standards.

Prompt Chaining in Action:

Here’s an example of how you might use it to write a short story:

  1. Prompt 1: Write a sentence introducing a lone robot on a deserted planet.
  2. Prompt 2: Following from the previous sentence, describe the robot’s appearance and its current activity.
  3. Prompt 3: Based on the previous prompts, introduce a new element to the story: a strange object the robot discovers.
  4. Prompt 4: Describe the robot’s investigation of the object, hinting at its potential purpose.

By chaining these prompts, you can guide the LLM to create a more engaging story than if you provided a single, broad prompt.


Prompt chaining opens a new chapter in AI interaction. By understanding its benefits, challenges, and best practices, you can unlock the true power of LLMs and achieve remarkable results in various fields. So, unleash your creativity, experiment with prompt chaining, and witness the magic of AI unfold!

About Lance Lingerfelt

Lance Lingerfelt Profile Photo

Lance Lingerfelt is an M365 Specialist and Evangelist with over 20 years of experience in the Information Technology field. Having worked in enterprise environments to small businesses, he is able to adapt and provide the best IT Training and Consultation possible. With a focus on AI, the M365 Stack, and Healthcare, he continues to give back to the community with training, public speaking events, and this blog.

Get E-Mail Updates
I agree to have my personal information transfered to MailChimp ( more information )
Want to know when I post new content? Sign up here to get an email when I do post!
I despise spam. Your email address will not be sold or shared with anyone else.