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?
I often use AI to create Macros or complex formulas for excel workbooks. I may run the provided code and wind up with an unexpected outcome. I am able to use that unexpected outcome to refine the initial request and find holes in the code that was created. Saving Changes...
Refining a prompt transforms vague, generic outputs into precise, actionable results. For example, changing "create a project plan" to "generate a 2-week agile sprint plan for a mobile app launch, including user stories, dependencies, and risk mitigation steps" yields a detailed, directly usable framework instead of a high-level template. Specificity in context, format, and constraints guides the AI to deliver higher-quality, relevant content. Saving Changes...
It will be always an interactive process to get to the more useful output. As much we work on prompt refinement as much result we will have. Saving Changes...
As a cognitive project manager in ML and AI, I’ve witnessed how refining prompts can significantly impact project success. It’s not just tweaking language—it’s about strategically crafting inputs to guide models toward outputs that are relevant, actionable, and aligned with stakeholder goals. In one project, we enhanced model performance by incorporating user personas and emotional tone into the prompt, transforming the output from generic summaries to tailored insights that truly resonated with our audience. In this sense, prompt engineering becomes a form of cognitive design—where understanding human intent and translating it into precise, machine-readable instructions is crucial for achieving impactful results.
Saving Changes...
DEYAA ABOUHASSANDARALRIYADHAl Majmaa / King Fahd District, 01, Saudi Arabia
Jun 21, 2024 7:28 AM
Replying to Sergio Luis Conte
...
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).
thanks for sharing that method Saving Changes...
Bandar BahlulManagement| Logistics ManagementRiyadh, 1, Saudi Arabia
Impact of Prompt Refinement on Output Quality:
From vague to precise
When prompts are general or unclear, the results are often superficial or not useful. But once the request is clarified, the output becomes more focused and professional.
Contextual customization
Adding details like the domain (e.g., customs clearance, project management, marketing) helps the system provide more relevant and in-depth responses.
Reducing repetition and errors
Improved prompts reduce the chances of inaccurate or repetitive information, increasing the reliability of the results.
Achieving specific goals
When prompts are crafted based on a clear objective (e.g., preparing a report, analyzing a problem, drafting a message), AI delivers results that are much closer to what you actually need.
Practical Example:
If you ask the system, “Help me create a project plan,” you’ll get a generic template.
But if you say, “Help me create a project plan for developing an electronic customs clearance system in Saudi Arabia within 3 months,” you’ll get a detailed and context-appropriate plan.
Saving Changes...
Gopal AsthanaService Coordinator and Incident Manager| KyndrylNoida, India
Honestly, I’ve seen the difference feel almost magical at times. A vague or rushed prompt usually gives you something generic, but when you take a step back, add just the right context, and guide the model with clarity, the output can go from “usable” to “wow, this is exactly what I needed.” Even small tweaks - like shifting tone, framing the role of the AI, or narrowing the scope - can completely change the depth and relevance of the response. It’s a bit like tuning a radio; once you hit the right frequency, everything comes through crystal clear. Saving Changes...
From my experience and understanding of the training, the difference between a vague, unrefined prompt and a refined chaining one is the difference between a rough draft and a finished product.
It's not just a subtle improvement; it's a fundamental change in the quality, relevance, and usefulness of the output... Saving Changes...
Ronald CairoSpecialist Engineer| Ministry of Housing, Construction and Sanitation of PeruElkridge, MD, United States
In a very particular way, I ask the AI to improve its results, after receiving the new answer, I ask again to improve the result, changing a new question with the two previous changes and after the answer, I ask again to improve its answer and reduce it in a simple, clear and easy to digest language and always cite the sources. Saving Changes...
PMI Team Member
