Project Management

Home / Discussions / Project Management Central /

When using AI systems, what are some best practices for ensuring the results you receive are accurate, relevant, and aligned with your original goals?

Artificial Intelligence

Sarah Philbrick

PMI Team Member

Director, Learning Design & Development| PMI Asheville, 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?

Posted: Jun 6, 2024 1:53 PM

Sort By:

< 1 ... 181 182 183 184 185 186 187 188 189 190 191 ... 208 >

Sibeti Mboo Zambia

Assigning a character or role to the AI in the prompt, providing real examples or documents as well as specific outputs to ensure the AI responses are aligned to the task. Additionally, refining the prompt and evaluating each response

Posted: May 17, 2026 1:29 AM

luis cantu Clearfield, Ut, United States

I've only used AI automation and assistance so far. The way I make sure it's accurate is by re-feeding it's output and using a persona prompt to audit the work. I then take the audited feedback and tailor to my need. I've found that the information it provides is helpful for my blindspots, but ultimately I assess if the risk can be accepted or not.

Posted: May 17, 2026 3:29 PM

Narayanan Ramasubramanian Chennai, TN, India

In your prompt make sure you :
Provide clear context
Provide details with specific instructions
Use RTF or CREATE Formula
Provide clear output format / requirement

Posted: May 18, 2026 12:39 PM

Shawn Robison Program Manager| Johnson & Johnson Fort Worth, Tx, United States

Some good recommendations in this thread. My $0.02...be clear about the role of AI in what you're asking, be specific about what inputs you want (or don't want) it to use, and be clear about the output you're looking for including the intended audience. Once you find something that meets your need turn it into a skill.

I've had good success with that approach and have managed to automate hours of administrative effort out of my week.

Posted: May 18, 2026 1:15 PM

Pradeep Shukla Continuous Improvement & RPA Lead| Bunge India Mohali, PB, India

Define clear goals and give specific prompts

Verify outputs with trusted sources

Refine prompts through iteration

Use critical thinking (don’t blindly trust AI)

Avoid sharing sensitive data

Review results with team when needed

Posted: May 18, 2026 2:05 PM

Pradeep Shukla Continuous Improvement & RPA Lead| Bunge India Mohali, PB, India

Define clear goals and give specific prompts

Verify outputs with trusted sources

Refine prompts through iteration

Use critical thinking (don’t blindly trust AI)

Avoid sharing sensitive data

Review results with team when needed

Posted: May 18, 2026 2:05 PM

Anonymous

Give AI specific examples to work with. Be as specific as possible regarding expected output and format. I have even given AI our brand materials so that presentations are consistent with branding.

Posted: May 19, 2026 3:42 PM

Chloe Pichon Global Liaison and Vendor Management Lead| Swedish Orphan Biovitrum Zurich, Switzerland

Being methodic and precise in what we ask (by using some of the RTF or CREATE formulas, for instance) is key, but I think it also comes from a good preparation from the project manager ahead of prompting, in order to clarify what goal they have and output they expect to receive from AI. Once this is clear, it gets easier to place a clear request to AI. And the more we practice, the better we all get, both the GenAI and the person using it.

Posted: May 20, 2026 8:24 AM

Anonymous

So far, I've preferred to keep prompts very general and not organization-specific. Mostly it's been beneficial with tools like using Co-Pilot in Excel or Power BI, where you can search and prompt different queries that haven''t been possible before.
Accuracy is key and when working with formulas and spreadsheets and AI, it's still been worth spot checking for accuracy.

Posted: May 21, 2026 4:54 PM

Ing Christian Nyame Senior Manager, Network Operations| MTN GHANA Accra, Ghana

Jun 08, 2024 1:37 PM

Replying to Keith Novak

...

Like with any new tool, you need to test the results before you scale up.

Think about if you were to manually model a very complex problem in a spreadsheet. You don't build all the links and formulas first and then evaluate your final output. You build and test sections of the bigger solution first and then add on layers once you have validated the functionality.

the core principle is treat AI as a “decision support system” and not a source of truth

Posted: May 22, 2026 11:00 AM

< 1 ... 181 182 183 184 185 186 187 188 189 190 191 ... 208 >

Please login or join to reply

When using AI systems, what are some best practices for ensuring the results you receive are accurate, relevant, and aligned with your original goals?

Sponsors

Newsletters