Transforming Ideas into Action: How AI Turns Concepts into Fully Planned Projects

Written By Dana Coffey

Most project managers begin with an “analog” approach when turning initial project ideas into documentation and plans. In my experience, many projects start with senior leadership mentioning an idea, identifying a problem, or suggesting an improvement. At this early stage, I often know very little. Discovery and discussion play a crucial role before drafting a Project Charter or Scope Statement.

Many project managers rely on templates to streamline this process. These templates often build on past experiences, incorporating common elements from similar projects. For example, in my current role as an IT project manager, I use templates that include common risks, requirements, and costs typical of custom business software projects.

Recently, I’ve been exploring Artificial Intelligence and Prompt Engineering, particularly how these tools can accelerate the artifact creation process. After months of experimentation, I decided to put my knowledge to the test using PMI Infinity. (PMI Infinity comes free with a PMI membership, and a free 30-day trial is available here: https://www.pmi.org/infinity). This article will explore the process of getting things started with very little information and generating a suite of project artifacts needed to move forward.

Note: while I used PMI Infinity for this exercise, it can be replicated using other tools; my top 3 recommendations are Chat GPT, Claude, and Perplexity.

Prompt Formulas and Patterns Used in This Exercise

AI Prompt Formulas Featured

Two key prompt structures that enhance interaction with AI for generating project artifacts are RTF (Role, Task, Format) and CREATE (Character, Request, Examples, Adjustments, Type, Extras).

  • The RTF formula is an efficient and structured way to guide AI by defining three essential elements: the Role that the AI should assume, the Task it needs to complete, and the Format in which the response should be delivered. This structure ensures that responses are targeted and organized according to specific needs. For example, when generating a project charter, you might instruct the AI to take on the role of a "senior project manager" (Role), request it to "outline the key components of a project charter" (Task), and ask for it to be presented as "a bullet-point list" (Format). This formula is especially useful for ensuring clarity and precision in responses.

  • The CREATE formula takes this a step further by adding more nuanced elements to the prompt structure. It includes Character, where the AI is given a specific persona or role to play; Request, which clearly outlines the task at hand; Examples, providing guidance on the desired output; Adjustments, allowing for modifications in the response; Type, specifying the format of the output; and Extras, offering additional context or information. For instance, when drafting a project scope document, you might ask the AI to act as "an IT consultant" (Character), "draft a project scope statement" (Request), provide examples of scope limitations (Examples), adjust to focus on a software project (Adjustments), present it as a "detailed plan" (Type), and include "specific risks and assumptions" (Extras). This method ensures comprehensive and well-structured responses, tailored to project-specific needs.

I generally use a mixture of these two.

For more on prompt formulas, visit https://fvivas.com/en/19-formulas-and-prompt-structures-for-chatgpt/

AI Prompting Patterns Featured

Several advanced AI prompting patterns enhance the process of generating project artifacts, including flipped interaction, the explainer bundle, and prompt chaining. You’ll see all three used to some degree in this exercise

  • Persona pattern allows the AI to take on the voice and perspective of specific stakeholders or experts, ensuring the output aligns with the needs and concerns of those personas. For instance, asking the AI to "act as a CFO" can help tailor responses to focus on financial priorities, helping to create documentation that resonates with different project stakeholders.

  • Flipped interaction, instead of the project manager giving direct instructions, the AI takes the lead by asking the user a series of guiding questions. This approach helps fill in knowledge gaps and encourages the project manager to think through key details that may have been overlooked.

  • The explainer bundle pattern is used to generate detailed explanations of complex topics, breaking them down into simpler, more digestible parts. This is especially useful when drafting project documents that need to communicate technical information clearly to non-technical stakeholders.

  • Prompt chaining involves linking multiple prompts together, allowing for a series of interactions where each response builds on the previous one. This pattern is ideal for iterative tasks, such as creating comprehensive project plans, where one prompt may generate an initial outline, and subsequent prompts refine each section in detail. Together, these patterns create a dynamic, flexible, and highly efficient approach to generating project documentation with AI.

For more on prompt patterns, visit https://www.vanderbilt.edu/generative-ai/prompt-patterns/.

The Idea and the Process

Things get a little hypothetical now; "The characters and events depicted in this program are fictitious. Any similarity to actual persons, living or dead, or to actual events, is purely coincidental."

So having lunch one day, the CEO, Bob Brown approaches me with a new idea he has. He’d like a special dashboard specifically for the Sales Team to see all the things they must tend to each day. They’d like to have a section for company Sales KPIs, their own individual performance, a customizable set of personal links, SharePoint integration, Office 365 Integration, and a few other integrations such as with SAP S4 Hana and Tableau reports. The sales team really doesn’t have time or inclination to go through a long discovery process, so I start with just this little bit and go over to PMI Infinity.

Context and Persona.

First, I instruct the AI Engine to use my voice with the following prompt:

I am a PMP Certified Project manager who has 30 years’ experience in Information Technology and software development. Before becoming a Project Manager, I was on the engineering team and developed apps for 25 years before switching careers. Act as me for the duration of this conversation.

Flipped Interaction

I use the paragraph above where I have a small description of my starting place and paste that as a prompt into PMI Infinity. I then ask it to ask ME all the questions it needs to ask to take my idea and turn it into a project charter and scope document.

The Questions

To follow are the questions that PMI Infinity asks me with my answers in red.

To generate a comprehensive project charter and scope statement, I'll need a bit more information. Here are some questions to help us get started:

  1. Project Goals and Objectives:

    • What are the primary goals and objectives of this dashboard project? To increase sales by 20% due to time regained from having a centralized information repository. To eliminate the daily sales meetings and streamline communication, to provide a centralized area just for the sales team and the way they operate

    • Are there any specific business problems or opportunities this dashboard aims to address? yes. the sales team is functioning in an individual way with each rep managing tasks and information in their own unique way. There isn’t a common view, common sop right now. They want a one stop place to see metrics, their sales performance etc and also manage their daily tasklists, answer emails, schedule zooms, and manage a custom list of hyperlinks that they usue regularly.

  2. Stakeholders:

    • Who are the key stakeholders for this project besides the Sales Team? Bob Brown - CEO and executive sponsor, Dana Coffey - project manager, Steve Spun - Systems Administrator, Allison Villeaux - Database Administratior, Jeffrey Tea - Development Lead

    • Are there any specific individuals or departments that should be involved or consulted? John Staccato and Mark Hanson - both VPs of sales

  3. Requirements and Features:

    • Can you provide more details on the specific KPIs and performance metrics that need to be included? Sales performance and margins, sales rep targets, integration with tableau kpi dashboard

    • What kind of customization options are expected for the personal links section? they want to keep a master list and be able to add/edit/delete links for the whole group; anybody can do this

    • Are there any specific requirements for the SharePoint and Office 365 integrations? They just want web parts on the site that allow them to connect to SharePoint document repositories

    • What specific data or reports from SAP S4 Hana and Tableau need to be integrated? existing tables and views

  4. Constraints and Assumptions:

    • Are there any budgetary constraints or timelines we need to be aware of? We need it completed by July 2025.

    • Are there any assumptions we should consider regarding the availability of data or systems? Use an assumption list typical of similar projects. (see what I did there? This will help stubb out some things that I might not consider and I can delete those I do not need; I use this technique a lot as you will see.)

  5. Risks and Dependencies:

    • Are there any known risks or dependencies that could impact the project? Start with a list of risks and dependencies typical of similar projects and add the following: low adoption risk, risk of scope creep

    • Are there any other ongoing projects or initiatives that might affect this dashboard project? We have two projects going that will use about half the resources that will also be assigned to this project.

  6. Success Criteria:

    • How will the success of this project be measured? Increased sales margins, fewer meetings for the sales team, and anything else that other simlar projects have used.

    • Are there any specific deliverables or milestones that need to be achieved? yes Milestones are Kickoff Nov 1, 2024, Prototype Review Jan 1, 2025, Beta 4/30/2025, Production July 1, 2025

  7. Resources:

    • What resources (e.g., team members, tools, software) are available for this project? Sql Server, Windows Server, .NET, REST

    • Are there any specific skill sets required for the team working on this project? varies - use examples from similar projects.

I paste all that back in and give the following prompt:

Use my answers and examples from other similar projects to create a project Chater and Scope Statement

The Generated Scope Statement and Project Chater Documents

Building on What We Have and Generating More Artifacts

Using prompt chaining, we’re going to now have PMI Infinity use what it knows so far to build the other artifacts we need. Prompt chaining is basically breaking a prompt down into many steps and telling the ai to use the step before it to answer the current prompt. Have a look at https://www.datacamp.com/tutorial/prompt-chaining-llm for a tutorial on how to implement this method.

Stakeholder Artifacts

Beginning with Stakeholders, I ask PMI Infinity to generate all the artifacts I need

Using this project charter and scope statement, along with what you know about from similar projects, please generate all the artifacts I need for Stakeholder Management, including a RACI Chart in tabular format

Artifacts Generated:

Requirements Artifacts

My prompt:

Please Use the information we have gathered so far and generate the necessary requirements management artifacts. Where my information is lacking, please use information from other similar projects.

Artifacts Generated


Risk Management Artifacts

To prompt PMI Infinity to create the Risk related documents, I prompt it similarly to what I did with Stakeholders and Requirements.

Please take all the information from these documents we’ve created so far and mix that with samples of other similar projects and generate all the necessary Risk Management Artifacts

Artifacts Generated

Project Plan and WBS

As with previous prompts I asked the AI to use what we’ve created and combine it with similar projects to come up with a project plan and a Work Breakdown Structure (WBS)

Artifacts Generated:

Schedule and Budget

Again, using the same methods as before, I prompt the AI to create the schedule and budget artifacts.

Quality Management and Test Plans

I use the same method as we’ve used for the last few sections to generate Quality Management artifacts.

We’ll use the same method to generate the last remaining artifacts below:

Communications Management Plan

Resource Management Plan

Procurement Management Plan

Stakeholder Engagement Plan

Change Management Plan

Issue Log

Lessons Learned Register

Project Closure Report

Conclusion

The set of artifacts created is a very good first draft that I can now use to extrapolate and drill into what components will apply to my specific project. I usually use these as starting points, and I make them better by adding detail and context as that information comes to light.

Leveraging AI tools like PMI Infinity in project management can significantly streamline the process of creating essential project artifacts, even with limited initial information. This approach demonstrates how AI can serve as a powerful assistant, helping project managers to:

  1. Rapidly generate comprehensive project documentation

  2. Uncover potential risks and considerations that might otherwise be overlooked

  3. Standardize project planning processes across different initiatives

By utilizing prompt formulas such as RTF and CREATE, along with advanced prompting patterns like flipped interaction and prompt chaining, project managers can harness AI to produce detailed, tailored project artifacts efficiently. This not only saves time but also ensures a more thorough and consistent approach to project initiation.

While AI tools provide valuable assistance, it's crucial to remember that they should complement, not replace, human expertise and judgment. The project manager's role remains essential in refining AI-generated content, ensuring alignment with organizational goals, and incorporating nuanced understanding of stakeholder needs.

As AI technology continues to evolve, its integration into project management practices will likely deepen, offering even more sophisticated support for project planning and execution. Project managers who embrace these tools and develop skills in effective AI prompting will be well-positioned to lead more efficient, innovative, and successful projects in the future.

Since I’m still learning, I’d welcome any comments below that will help me build my skills in this area!

Sources

Dana Coffey
Previous

Leading Without Chaos: Avoiding Common Pitfalls That Undermine IT Success
Next

The Critical Role of Emotional Intelligence in Hybrid Project Management