Project Management Central
|
Please login or join to subscribe to this thread
|
|||
|
|||
always start from basics and be as simple as possible, provide clear instructions and build on your request to bring out what you really are looking at .once AI system gives responses validate and continue until you get what is usable.
Before implementing any AI solution, it is essential to set clear and measurable goals.
1- Data collection and analysis AI is powered by data to generate insights and make intelligent decisions. It is therefore essential to collect and analyze relevant data effectively. This means having the right systems in place to capture and store relevant data on project performance. 2-Automation of processes One of the most obvious advantages of AI in project management is its ability to automate repetitive and tedious processes. This allows project teams to focus on more strategic, high-value tasks and on identifying new business opportunities. 
Here’s how to ensure your AI systems deliver accurate, relevant, and goal-aligned results based on my experience as a Defense Industry Project Manager:
- First, do the research and literature study. This helps you understand the current state of AI and its applications in your field. Identify the best documentation to provide context. For example, in one of our projects, we started by reviewing recent studies on AI in military simulations. - Next, write down manually the workflow you will follow. This typically involves creating a chain or tree of thought, setting up a chain of feedback, and asking the model about worn-out ideas versus the valuable insights. Documenting this helps you stay organized. In one instance, we mapped out the entire AI integration process, which included data collection, model training, and validation steps. - Keep track of all interactions in a document. This is crucial for reviewing and improving the AI system. We maintained detailed logs of every interaction with our AI models, which helped us track progress and pinpoint issues quickly. - Evaluate answers iteratively. Don’t just settle for the first result the AI gives you. Review and refine the answers. We often found that initial AI outputs needed several rounds of evaluation and tweaking to meet our standards. - Combine answers and jump between steps as needed. Flexibility is key. Sometimes, you’ll need to revisit previous steps based on new information or results. In our projects, we frequently revisited our initial research and workflow documentation to adjust our approach based on the AI’s performance and feedback from stakeholders. These are the steps I follow to enhance the effectiveness and reliability of AI systems, ensuring I somehow meet my objectives and adapt to evolving needs.
Praveen Narayanan M
Bangalore, KA, India
Talking about generative AI all related to technology has almost not impact with relation to all related to non-technological roles and activities. What you stated about accuracy and things like that are easy to implement because there are a lot inside disciplines like statistics. Most of them to make things "a priori" to prevent instead of cure. Few organizations taking into account that when generative AI environments are put in place almost a new business unit has to be created where roles like lawyers, linguistic, diversity and inclusion specialist must be hire to help on put it in place.
Define very well your inputs, simplify your questions as much as possible, give feedback to the AI answers, provide AI with examples you already have, review with peers the outcomes of your discussion.
Clearly define your objectives: Before implementing any AI solution, make sure you have a clear vision of what you want to achieve. 
The usage of the CREATE model combined with refining tecniques can lead to an output aligned with the expectations in terms of accuracy, relevance and alignment.
I think a lot of trials and experiments are the way to obtain good results.
Ecue-Mathe Mensah
Benin
Refinement of your prompts is one of the best practice for ensuring the results you receive are accurate, relevant, and aligned with your original goals. By giving Feedback to the AI system, reworking on your prompt with specificity and details about what is expected from the AI response. The more concise and clear is our prompt, more accurate is the output of AI. 
PRANAV SHUKLA
Associate VP - Customer Support (Projects, Vendor Management, Customer Success)| TSI PLC.
Mumbai, Maharashtra, India
specific , clear instructions with examples and reiterate till we get the perfect answer.
Renzo Morante
Wake Forest, Nc, USA
Good question. Definitely how detailed you are with your prompt will have a huge influence with the outcome. Re-read your prompt multi times and refine it as needed. After as response is provided you need to ask for a summary of the resources that were used on the response. That will validate how meaningful is the response.
|
|||
PMI Team Member
Please login or join to reply
|
That's the true spirit of Christmas; people being helped by people other than me. - Jerry Seinfeld |
