Director, Learning Design & Development| PMIAsheville, NC, United States
With Generative AI, iteratively refining and optimizing prompts can lead to better AI-generated results. This may involve adjusting the specificity or clarity of the prompt to increase relevance and accuracy of results.
What examples do you have of how improving a prompt drastically changed the output quality? What specific changes did you make that led to the improvement?
Generic prompts tend to produce surface-level responses, whereas well-structured, domain-specific inputs directly impacts the relevance, clarity, and strategic value of the output, while using terms like clinical decision support, risk stratification, or workflow optimization—generate insights that align with operational and clinical priorities. For example, refining a prompt from “Explain AI in healthcare” to “Outline how GenAI can optimize hospital readmission risk prediction models” resulted in actionable, implementation-ready recommendations. Precision in prompt design is now a core competency for leveraging GenAI effectively at the executive and project leadership level. Saving Changes...
By using prompt Chaining we can write and break the complex requirements into small sub tasks to get better output quality .
Kindly aware of the input , be specific in your requirements to get the targeted output , I like to thanks all the members in dicussion forum for sharing your experience , i would like to thank PMI team for this wonderful course ( Prompt Engg fo PM ) to rule AI world ahead of us . Saving Changes...
The more you refine the prompt, the more accurate and desired output you get. It's like training someone or an individual on how to answer a particular question over time; the person becomes better at it and so will output a better result. Saving Changes...
Amanda ThomasProgram Manager| Mary Washington HealthcareFredericksburg, Virginia, United States
This was a really good section that stressed the importance that AI is an iterative process and needs information distilled, examples and gentle correction to ensure that you get the output you want. Essentially, you must be a good steward and good manager of the AI to be able to get the most out of it. Saving Changes...
There are framewoks to create prompt. This is part of the Prompt Desing discipline. Those that gave me and the initiatives where I was included are:R-T-F (Role-Task-Format), T-A-G (Task, action, goal), B-A-B (Before, after, bridge), C-A-R-E (context, action, result, example), R-I-S-E (role, input, steps, expectations).
I 100% agree with using these frameworks. But I'll also add that if you are a beginner looking to get good information from a chat based LLM, a great place to start is simply describing your issue/task then asking the LLM what kind of prompt it would give itself and ask what additional information would lead to a good answer. Modern LLMs will literally tell you the best way to prompt them.
These frameworks were extremely important when LLM models had short context windows and had a stronger penchant for hallucination, with the release of ChatGPT 3o (and now 5.0) and Claude 4.0, these frameworks are still an important best practice but not as necessary as they once were because you can iterate and add context to form an answer through a "discussion" in the chat.
In summary, these frameworks are best practice, but if you are just starting out, don't let them scare you either. You can iterate your answer through a context window of 1 million tokens with some of the leading models.
Being concise and specific helps the AI to give some valuable answers. It also learns with time as you ask further questions.
Refinement of the prompting always helps in better results. Even if we use the chain of commands, we get the finest results with refinement prompting. Saving Changes...
Anonymous
My experiences have been very human like. I just mentioned the output parts I don't like or isn't easy to use and how it can be rectified. Sometimes AI makes improvement suggestions as well. Saving Changes...
Iteratively refining a prompt for boiler tube analysis from a vague query like "analyze boiler tubes" to a more specific one drastically improved the output quality. By adding details about the desired analysis, including metallurgical failure analysis of a waterwall tube and specifying creep damage and corrosion mechanisms, the output became more highly relevant and accurate. I am excited to be part of this community to collaborate and improve. Thank you all. Saving Changes...
I was recently in charge of redesigning a work procedures manual in an organization that formulates research and executes projects at a national level; there I used generative AI to contrast an improvement provided to address better performance in the execution of projects in the client organization, I really had to make quite a few refinements to achieve a product close to my expectations, since I had to incorporate specific tasks into the LLM in different roles that were part of the project beneficiaries, in the end I obtained substantially improved products, but not before investing time in the refinements and after making several iterations improving the prompts Saving Changes...
Haresh ThevathasanSr. IT Consultant| Public Consulting GroupCa, United States
By clearly stating a need and clearly defining outputs AI will generate a fairly good output of your request. Examples of previous templates are key. However please keep in mind that you are ultimately the PM that needs to define the final deliverable. Saving Changes...
PMI Team Member
