How to Create a WBS from Scope of Work Using Claude, ChatGPT, and Plannex AI
Artificial Intelligence is becoming a powerful assistant for planning engineers, especially when working with repetitive or structure-based tasks such as creating a Work Breakdown Structure, preparing activity lists, organizing procurement packages, or generating Primavera P6-ready data.
In this tutorial, we test how to create a WBS from a scope of work using three different AI workflows:
Claude AI, ChatGPT/OpenAI, and Plannex AI.
The goal is simple: take a project scope of work and convert it into a clear, structured, and Primavera-ready WBS that can later be exported as an XER file and imported into Primavera P6.
Project Example: 18-Floor Hotel Tower
For this example, the project used is an 18-floor five-star hotel tower.
The project includes a basement, ground floor, two mezzanine floors, and fourteen typical floors. The scope also includes structural works, architectural works, MEP systems, procurement requirements, and floor-by-floor project breakdown information.
This type of project is a good example because it contains multiple disciplines and levels of detail, which makes it useful for testing how well each AI tool can understand and organize project information.
Why the Prompt Matters
Before using any AI tool, the most important step is preparing a strong prompt.
The prompt used in this workflow asks the AI to act as a senior planning engineer with experience in WBS development and construction project planning.
The prompt also includes clear instructions such as:
- Follow the required WBS structure.
- Do not add, remove, rename, or reorder columns.
- Each row should represent one WBS item only.
- Organize the WBS logically based on the project scope.
- Keep the output compatible with Plannex and Primavera workflows.
These instructions are important because AI tools can easily generate useful information, but without strict formatting rules, the output may not be ready for direct use in Excel, Plannex, or Primavera P6.
Testing the Prompt with Claude AI
The first test was done using Claude AI inside Excel.
The project scope was attached, the prompt was added, and Claude generated a WBS based on the provided scope of work. The output showed a good understanding of the project structure and created a logical breakdown based on major project areas, disciplines, and work packages.
Claude was also tested again with additional instructions to improve the procurement breakdown. This helped generate more detailed procurement-related WBS items, which is important for real construction planning workflows.
Testing the Same Prompt with ChatGPT
The same prompt and project scope were then tested using ChatGPT/OpenAI inside Excel.
ChatGPT also generated a complete WBS structure, but the result was reviewed and compared against Claude’s output. The comparison focused on the level of detail, discipline breakdown, procurement structure, and how well the generated WBS could be used later inside Primavera P6.
Using the same prompt for both Claude and ChatGPT makes the comparison more realistic because both tools are working from the same project data and instructions.
Creating the WBS with Plannex AI
After testing Claude and ChatGPT, the same workflow was tested using Plannex AI.
Plannex includes a ready option to create or build a WBS, which makes the process easier for planning engineers who want a structured output inside Excel.
The WBS generated by Plannex AI was reviewed and compared with the Claude and ChatGPT outputs. Since Plannex is designed specifically for planning and project controls workflows, the output was more directly aligned with the next steps, especially exporting to XER and importing into Primavera P6.
Exporting the WBS to XER
Once the WBS structures were generated, the next step was exporting each result to an XER file.
This was done for:
- Claude-generated WBS
- ChatGPT-generated WBS
- Plannex AI-generated WBS
The purpose of exporting to XER is to make the AI-generated WBS usable directly inside Primavera P6.
This step is very useful because it connects AI-generated planning data with real scheduling software, reducing manual work and saving time during schedule development.
Importing the WBS into Primavera P6
After exporting the XER files, the next step was importing them into Primavera P6.
Inside Primavera, the WBS structures were reviewed and compared side by side. This helped identify how each AI tool handled the project hierarchy, discipline breakdown, procurement items, and floor-based structure.
This comparison is important because a WBS may look good in Excel, but the real test is whether it works properly inside Primavera P6.
Key Lessons from the Test
This workflow shows that AI can support planning engineers in creating a WBS faster, but the quality of the result depends heavily on the prompt, the project scope, and the tool being used.
Claude and ChatGPT can both generate useful WBS outputs, especially when given clear instructions. However, Plannex AI provides a more planning-focused workflow because it is connected to Excel-based planning tools and XER export features.
For planning engineers, this means AI can be used not only for ideas or text generation, but also for practical project controls tasks such as WBS creation, activity list preparation, BOQ structuring, resource planning, and Primavera-ready data generation.
