Director, Learning Design & Development| PMIAsheville, NC, United States
Validating and checking outputs is critical when working with AI systems like Generative AI. Such validation approaches may include establishing clear criteria, implementing strong testing protocols, and continuous refinement.
In your experience with AI, what are some best practices for ensuring the results you receive are accurate, relevant, and aligned with your original goals?
Julien PerretFull Time Lecturer| HEC MontrealLongueuil, Quebec, Canada
GenAI can be a wonderful tool and should be considered as such, a tool, that can bring undesired outcome if missused.
First, is think it's important to a reflect a bit before jumping on a GenAI tool to solve a problem. The best use case for GenAI is still automation and ask the tool to build a model to automate a process for which we know clearly the desired outcomme. It could be a model in Excel, an HR process, Project planning and optimisation, and so on.
The use of GenAI get exponential results if not use for a single outcome but when ask to propose a way to automate the proper generation of the outcome instead ( I get that advice from a conference organised by Google for startup).
So, it's better to ask the tool to help build a model to get the right instead of asking what is the right answer.
Personnaly, what I do regularly is copying the original prompt with the first outcomme and then highlighting the biais, errors or subject not covered by the answer and asking to reoptimize the answer. It reinforce the quality of the answer and the AI model as well (if you allow to share your prompt with the plateform editor you are using).
Thanks all for the knowledge sharing! Saving Changes...
Riaz MohammedProject Management Unit Head| Al Kuhaimi Metal IndustriesDammam, Saudi Arabia
Hi...Please explain in more details about PROMPT ENGINEERING. Saving Changes...
Manohar Lal DhimarOperations Head| SINAI Healthcare Private LimitedBhopal, India
As project managers begin leveraging AI tools for analysis, reporting, and decision support, it’s essential to ensure that the outputs we receive are accurate, relevant, and aligned with project goals. Based on my experience and learning, a few best practices stand out:
Start with a Clear Objective:
Define what you want the AI to accomplish. A well-framed question or prompt—linked directly to project objectives—significantly improves the quality of results.
Provide Context and Constraints:
AI performs best when it understands the background and boundaries of the problem. Include project details, assumptions, and the desired format of the output to keep the response aligned.
Verify and Validate Outputs:
Treat AI-generated information as a first draft, not a final answer. Cross-check data, validate facts, and confirm consistency with organizational standards or PMI frameworks.
Use Iterative Refinement:
Refine prompts progressively. Asking follow-up questions or giving feedback helps AI systems deliver more targeted and actionable results.
Maintain Ethical and Confidential Standards:
Avoid sharing sensitive project data with external tools. Ensure compliance with data privacy and intellectual property policies.
Balance AI Insights with Human Judgment:
AI can analyze patterns and summarize information rapidly, but strategic thinking and decision-making must remain human-led. The best results come when AI augments, not replaces, human expertise.
In essence, AI is a powerful collaborator—its value depends on how clearly we communicate our goals, interpret its results, and apply professional judgment.
Saving Changes...
Manohar Lal DhimarOperations Head| SINAI Healthcare Private LimitedBhopal, India
As project managers begin leveraging AI tools for analysis, reporting, and decision support, it’s essential to ensure that the outputs we receive are accurate, relevant, and aligned with project goals. Based on my experience and learning, a few best practices stand out:
Start with a Clear Objective:
Define what you want the AI to accomplish. A well-framed question or prompt—linked directly to project objectives—significantly improves the quality of results.
Provide Context and Constraints:
AI performs best when it understands the background and boundaries of the problem. Include project details, assumptions, and the desired format of the output to keep the response aligned.
Verify and Validate Outputs:
Treat AI-generated information as a first draft, not a final answer. Cross-check data, validate facts, and confirm consistency with organizational standards or PMI frameworks.
Use Iterative Refinement:
Refine prompts progressively. Asking follow-up questions or giving feedback helps AI systems deliver more targeted and actionable results.
Maintain Ethical and Confidential Standards:
Avoid sharing sensitive project data with external tools. Ensure compliance with data privacy and intellectual property policies.
Balance AI Insights with Human Judgment:
AI can analyze patterns and summarize information rapidly, but strategic thinking and decision-making must remain human-led. The best results come when AI augments, not replaces, human expertise.
In essence, AI is a powerful collaborator—its value depends on how clearly we communicate our goals, interpret its results, and apply professional judgment.
Saving Changes...
Manohar Lal DhimarOperations Head| SINAI Healthcare Private LimitedBhopal, India
As project managers begin leveraging AI tools for analysis, reporting, and decision support, it’s essential to ensure that the outputs we receive are accurate, relevant, and aligned with project goals. Based on my experience and learning, a few best practices stand out:
Start with a Clear Objective:
Define what you want the AI to accomplish. A well-framed question or prompt—linked directly to project objectives—significantly improves the quality of results.
Provide Context and Constraints:
AI performs best when it understands the background and boundaries of the problem. Include project details, assumptions, and the desired format of the output to keep the response aligned.
Verify and Validate Outputs:
Treat AI-generated information as a first draft, not a final answer. Cross-check data, validate facts, and confirm consistency with organizational standards or PMI frameworks.
Use Iterative Refinement:
Refine prompts progressively. Asking follow-up questions or giving feedback helps AI systems deliver more targeted and actionable results.
Maintain Ethical and Confidential Standards:
Avoid sharing sensitive project data with external tools. Ensure compliance with data privacy and intellectual property policies.
Balance AI Insights with Human Judgment:
AI can analyze patterns and summarize information rapidly, but strategic thinking and decision-making must remain human-led. The best results come when AI augments, not replaces, human expertise.
In essence, AI is a powerful collaborator—its value depends on how clearly we communicate our goals, interpret its results, and apply professional judgment. Saving Changes...
To me using AI is similar than making questions or speaking to somebody else. The more clear and detailed your inquiries are, giving enough context, avoiding assumptions, being emphatic, listening well and asking again if not clear yet, are the best approach to get the right answers. Saving Changes...
To get accurate and actionable results from the Gen AI tools, the most important thing is the clarity of problem or situation in the mind of the user. and secondly what kind of solution is needed is also needs to be cleared in the mind of user. After that comes the different techniques to give effective prompts to the Gen AI tools. Saving Changes...
PMI Team Member
