
The VIKTOR AI platform lets engineering teams build, share, and govern engineering automation apps. You can create apps that integrate with SCIA in minutes using VIKTOR's built-in LLM, then publish them to the rest of your organization through a centralized app store. Every app runs through a governed pipeline. It is thoroughly tested and made accessible to the right people. That means the app you built today can be shared and reused by others on their projects.
In this blog, you will automate the following SCIA workflows:
A parametric ring frame structure: to show how to generate design alternatives for complex geometry
A parametric concrete shear wall app: relevant for pre-designing and iterating on 2D elements
A foundation slab on soil springs template app: relevant for exploring design alternatives for slabs
The first automation creates a circular structural model made from ring beams and columns. It is a good starting point because it keeps the workflow close to what engineers usually need first: create the model, run the analysis, and bring the raw SCIA results back into the app. You can change the radius, number of columns, column height, section dimensions, and vertical load, then review the internal forces and reactions without rebuilding the model by hand each time. You can also use this as a template to create or prompt more complex geometry, which saves time and gives you a reusable starting point for comparing structural alternatives.
To create the first automation, use the prompt box below. It will redirect you to the App Builder, our engineering AI that will help you create the first version in the VIKTOR platform.
After App Builder finishes creating the app, you can download and install VIKTOR Desktop. The process is simple. The video below shows how to download VIKTOR Desktop, sign in, configure the worker, and start the connection.
After VIKTOR Desktop is online and set up, upload an ESA file template that matches the description in the prompt. For this example, you can download scia_template.esa. Then you should be able to run the app and get a parametric ring frame structure.
After the first two examples with frame elements, we move to a 2D element: a concrete shear wall. This type of automation is useful for quick sizing and early estimation, because you can change the wall length, height, thickness, material, and load directly from the interface. The app still returns the complete SCIA table, so you can inspect the raw 2D internal force output first. Then it adds a heatmap where you can choose whether to visualize n_x, n_y, or n_xy, which makes the results easier to read without losing access to the original table.
After App Builder creates the app, upload the ESA file template in the app. For this example, you can use scia_template.esa. Then run the workflow and you should be able to see the results.
Finally, after exploring internal stresses in 2D elements, we can use the same idea to visualize deformation. In this case, the app creates a foundation slab on soil springs, which is a typical use case when estimating footing slabs or slab foundations. Because the slab dimensions, thickness, and subgrade stiffness are inputs in the app, you can quickly compare different geotechnical assumptions and foundation sizes. Instead of only checking a table of displacements, the heatmap helps you see the deformation pattern and understand where the slab is moving the most.
After the App Builder creates the app, upload the ESA file template in the app. For this example, you can use scia_template.esa. Then run the workflow and you should be able to see the results.
In my experience working with the VIKTOR and SCIA integration, the App Builder gives you a strong first version that you can quickly review and improve. The following tips will help you validate the generated model in SCIA, give the App Builder more context about the expected results, and continue prompting until the workflow matches your engineering process.
The apps in these examples include two buttons to download viktor.xml and viktor.xml.def. Keep both files in the same folder, import the generated XML model into SCIA, and run the analysis there. This lets you review the geometry, sections, supports, loads, and results directly in SCIA before using the workflow on a project.
If something needs to change, return to the App Builder and describe the improvement. This creates a useful review loop between the generated app and the engineering model.
Open the I/O document named output in your ESA template and take a screenshot of the result table you want the app to read. Upload the screenshot to the App Builder with your next prompt. The image gives the App Builder more context about the table names, columns, load cases, and result hierarchy.
This is especially useful when the parser assumes that the ESA file contains several nested result sections and searches under the wrong parent group. Showing the actual I/O document helps the App Builder retrieve the intended table and return the correct results.
These practices make the first version easier to review and improve:
Start with a circular, circular hollow, or rectangular section that the integration can create directly. If you later use an existing profile such as IPE200, make sure it is available in your SCIA setup and use its exact SCIA name in the prompt.
Check the units in both the prompt and the generated model. Geometry commonly uses meters, forces use newtons, moments use newton meters, and angles use degrees. A quick human review prevents misunderstandings from reaching the final results.
Make sure the ESA template contains the I/O document named output and includes every result table the app needs. Without that document, the worker cannot return the expected results.
Review the SCIA integration documentation to see the supported model parts and available methods. Even when a simple prompt gets you far, the documentation can help you discover new ways to extend the workflow
With these three automations, you can create SCIA models, run analyses, inspect internal forces, review reactions, and visualize 2D results from one VIKTOR workflow. This saves time when you test alternatives and reduces rework when a similar model or result check appears in another project.
You can also package these automations into apps that can be shared, tested, improved, and scaled across your team in the VIKTOR platform.
Try one of the prompts with the SCIA template above using the App Builder, and adapt it to a workflow your team repeats often. You can start with VIKTOR here