Artificial intelligence is a common topic in CAD, but too often it’s talked about in big, abstract terms. To make the topic more concrete for Octave BricsCAD users, we asked two of our experts, Mark Tooley and Julie Keverian, to answer ten questions about how AI-driven tools work inside the software. From predictive commands to drawing cleanup and guided editing, they explain what problems these tools were designed to solve and where human expertise still plays a critical role.
Here are ten practical questions—and straight answers—about AI in BricsCAD.
When people hear “AI in CAD,” what does that mean inside BricsCAD? In plain language?
Mark Tooley: In simple terms, it’s all about reducing manual, repetitive tasks to automate, accelerate, or improve tasks traditionally done manually. Drawing Health is a good example, as you can run file clean ups to remove unnecessary detail and repair errors and reduce file size.
Julie Keverian: It’s really a faster way to accomplish a task, with some kind of automation. Think of this as “taking the drudgery out of drafting” and removing repetitive tasks.
What real user problem(s) kicked off the development of these AI-driven tools?
Tooley: The typical day of a designer working with CAD software consists of two types of tasks. On the one hand, they do high-value, creative activities, solving design problems, developing solutions for their customers, and so on—the fun stuff. But also, there’s a lot of repetitive, laborious tasks that inevitably have to be done alongside the more interesting tasks.
With BricsCAD, the “Aided” in CAD means helping designers get as much of the low value work done as quickly as possible by automating most repetitive tasks. This then gives them time for more of the high-value creative tasks and doing actual "design" work, which results in better, more innovative outcomes, and faster delivery of design work downstream. This could mean getting better products to clients quicker, or earlier fulfillment of contracts, essentially enabling the organization to get paid faster.
Keverian: Traditional CAD hasn’t changed for decades, meaning we’re drawing lines, arcs and circles today, just as we did 40 years ago. BricsCAD saw the opportunity to make CAD even more productive with our unique productivity tools, including Copy Guided, Blockify, and Drawing Health.
A lot of this focuses on removing repetitive work. What’s a small but constant task users don’t realize is slowing them down?
Tooley: CAD drafting can be heavy on menu selection. Having the ability to select from contextual menus on your mouse cursor is a great time saver. If you totaled up the time saved in mouse travel up and down the screen over a week, it adds up.
Keverian: CAD makes it easy to reuse drawings and any content they include without creating them from scratch the next time you need them. Content such as Dim styles, Text styles, page setups, blocks, etc. The key is to know where that content is and how to easily access it. Drawing Explorer, although it’s not an AI tool, makes it easy to manage that content and copy it between drawings.
How does BricsCAD’s AI learn from user behavior without sending data to the cloud?
Tooley: The AI Predict ribbon anticipates your next steps and recommends relevant commands based on your workflow. You can opt in to the BricsCAD AI Analytics Program to get personalized suggestions.
Keverian: It depends on the feature. Some AI capabilities in BricsCAD work entirely on your local machine. For example, QUAD command predictions use a machine learning model that ships with BricsCAD. The model analyzes the commands you use and adapts the suggestions it shows, helping the interface anticipate your next step. Because this model runs locally, it can adjust to your workflow without sending drawing data or usage information to the cloud.
Other AI-driven features, such as the AI Assist ribbon, require participation in the BricsCAD Analytics Program. In that case, usage data is shared with our servers to enable more personalized recommendations. In short, some AI features learn locally, while others rely on optional analytics, giving users control over how their data is used.
What’s the difference between traditional automation and what these tools are doing under the hood?
Tooley: Standard CAD automation typically involves batch file conversion using software tools, scripts, rules, or APIs to automatically generate, modify, or validate CAD models and drawings, reducing repetitive manual work and improving accuracy. While AI tools alleviate repetitive manual work, they focuses on patterns within your workflow. Through machine learning, it predicts what you are trying to do, adding smart behavior to the design process.
Keverian: Traditional automation could be regular blocks. So, instead of drawing a gear from scratch every time you need it, create a static block. It's ready when you need it, just insert the block into the drawing. Then take the static block and make it parametric, with various sizes. Create it once and use it many times.
Drawing health and cleanup come up a lot. Why is this such a big deal for performance and reliability?
Tooley: Having drawings with lightweight file sizes speeds up your workflow, and by removing unnecessary details, this ensures accuracy in the output data.
Keverian: As drawing size grows, performance can drop. If it takes longer to open a drawing, or to pan and zoom around the drawing, productivity also decreases and it takes longer to perform even simple tasks. Time is money. Also, let’s consider blocks. Say you have some duplicate block definitions, meaning Block A and Block B contain the same geometry but have different names. When you create a table based on how many times each block is inserted in the drawing, the count is off. There are two identical blocks, but with different names.
Guided editing tools adapt to context. What does “context” actually mean here in practice?
Tooley: The tools are enabled and display options to use when a specific type of geometry is selected. An example would be Copy Guided, as it automatically adapts copied geometry to its new context, aligning and adjusting elements for a seamless fit.
Keverian: Copy Guided will copy a door from one location to another, using the original selection as a guide for the placement of the copy. That means it could cut out the lines representing walls, at the door opening, or copy the door a specific distance from another wall. With Move Guided, the lines representing the walls will be "healed" or rejoined when the door is moved.
Predictive suggestions sound powerful, but also risky. How do you make sure they stay helpful, not annoying?
Tooley: Less is more, and one example would be the Quad. It’s a very useful tool that the user can customize to suit their work pace. The icon size can be modified, and the user can determine how quickly it appears, making it useful and not intrusive.
Keverian: They need to be easy and helpful, not annoying. The Quad remembers the most recent commands. Draw a rectangle, offset it, then select both rectangles to apply a hatch pattern. Now, if you draw another rectangle, the Offset and Hatch commands appear on the top row of my Quad.
In BIM workflows, where does AI add the most value today? Where does human judgment still matter?
Tooley: Propagate for detail automation is a value add as it replicates design details across your model, ensuring consistent connections and components while significantly reducing repetitive work. There will still be human interaction, ensuring client delivery is accurate in terms of final checks for consistency—whether this is title blocks, annotation or dimension styles, we are still required!
Keverian: In general, with all AI, human judgment absolutely still matters. Let the tools suggest the results, but allow us to accept or reject them.
If a user only tries one AI-driven feature this week, which one would you recommend, and why?
Tooley: Drawing Health, particularly if you are receiving data from third parties. You can optimize the data, ensuring you start with a good foundation before investing time editing an inefficient file set.
Keverian: I'd also say Drawing Health. We all have drawings we've used over and over. And cleaning up our drawings is not something we remember to do on a regular basis. Most people don't think about it until the drawing starts misbehaving. We continue to develop and improve our tools, adding functionality with new releases. If you haven’t tried Drawing Health or Blockify lately, take a look at them in BricsCAD V26.
AI that works with your workflow
AI in BricsCAD shows up in practical ways that improve everyday workflows. Whether it’s predicting the next command you’re likely to use, automatically adapting copied geometry to a new context, or cleaning up drawings so they run faster and more reliably, the goal is the same: reduce repetitive work so designers can focus on design.
At the same time, these tools are designed to support—not replace—human judgment. Designers remain in control of their models and drawings, choosing when to accept suggestions, automate tasks, or refine results.
As AI capabilities continue to evolve, BricsCAD’s focus remains on delivering practical intelligence with tools that fit naturally into existing workflows and make CAD work better for you and your projects.
Want to see these AI-driven tools in action? Explore the latest BricsCAD features and try them in your own workflow.
