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?
The best practices would be: Be precise and clear, use appropriate field terminology, provide the context explaining your requests, indicate the outcomes according with your expectations, refine and interact with the bot to check and improve its responses. Saving Changes...
I make sure I provide enough examples so that the system is clear on the output that I'm looking for.
Saving Changes...
Solimar LatercaGlobal IT Data-Driven Agile Leader| Capgemini2281, Brazil
Using AI effectively starts with intentional and structured interaction. Project Managers unlock the best results when they treat AI as a strategic partner and use advanced prompting methods such as Chain‑of‑Thought, Tree‑of‑Thought, ReAct, flipped interaction, and iterative refinement to increase clarity and insight. Applying structured frameworks like CREATE helps ensure accuracy by being highly specific, providing context and examples, guiding the model through logical steps, and defining clear roles and expectations. Embedding ethics, security, and governance is essential—requesting sources, enforcing compliance and data‑privacy rules, and asking the model to review and validate its own output. Above all, continuous iteration refines quality and alignment, turning AI into a reliable accelerator of value that strengthens decision‑making and project outcomes. Saving Changes...
LaWanda YoungProject Management| USAFSt. Clair Shores, MI, United States
Jun 12, 2024 1:31 AM
Replying to Jabin Geevarghese George
...
When using AI systems is very hard to set the precision or accuracy of the responses. I love bringing in the Agile mindset here pretty much imagine if you are mentoring someone you do a Q&A and based on the reponses of your Mentee you give the feedback so that Mentee can align his/her thoughts in the direction that we hint similarly review the AI responses and using our rationale judgement
1- Give Feedback to the AI system
2- Rework on your promp and be specific on what is expected
3- Keep it short and conscise, guage the responses and slowly we can tune the AI system in a way to get the best output
4- Now the Tech. Solution that comes in for accuracy is havig specific set of APIs that talk to real and accurate data sources or use 2-3 outputs of LLMs and then analyze and bring the best in output.
I agree. Utilizing the best prompt formula and adjusting prompts according to the output received. Saving Changes...
Prior to beginning this course and learning about prompt formulas, I found that I received the best responses when I provided as much context and details as possible. I would include thing such as the intended audience, desired tone, background of what I'm asking for and why it's being created and how brief or detailed I wanted the AI's response to be. Initially I was definitely using it like "Google" rather than an evolutionary tool that can do more than just tell me what a capitol of a state is! Saving Changes...
Providing the specific context in clear and consise way is essential.
and of course ensuring the responses are in line with your thought process and desired outcome This should be by validating every response with clear and continued refinement of the prompts. Saving Changes...
I would ask the AI system to justify its answer or cite sources. Besides; I would use the following verification techniques:
Compare with trusted documentation
Use multiple prompts
Use another AI or tool for comparison
Perform manual expert review
Saving Changes...
Anonymous
I am learning that AI is a tool with maximum capabilities given with direct and concise instructions. Using mechanisms like ReAct allow AI to know when to adjust criteria, and it will not always get everything correct at first glance, but providing feedback, explaining criteria, and being intentional with the tone of the prompt or valuable in getting the correct feedback. Saving Changes...
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.
I agree with this approach. Starting small and scaling over time is always helpful. Saving Changes...
PMI Team Member
