This is the second in a three-part series from Aaron Perry's NXT BLD 2026 presentation. Read part one.
We've seen countless examples of AI generating images and helping architects explore ideas. That's valuable. But ideation is only the beginning. The real opportunity is AI that acts inside the BIM model, improving data, geometry and deliverables automatically.
AI in AEC to ideate
AI in Qonic to execute
What automation in BIM should look like
Qonic has developed practical and valuable use cases that significantly enhance the utility and quality of its models. We build these automations based around these core principles:
- Finding blind spots and missing data
- Actioning fixes, improving the data quality
- Adding new geometry, or repairing existing
- Propagating fixes across your model where viable, engaging you if needed
- Meeting your model quality requirements
- Enabling you to benefit from third-party automations
Let’s talk about three examples being used in real projects today:
1. Data Quality: Don't Just Query Errors—Resolve 15.000 Items in Minutes
Where most AI tools help you discover a problem, Qonic is the only place you can action them at the same time. Through natural language processing, you or a colleague with no prior BIM software experience can action an outcome. Open a browser, ask a simple question: ‘how many doors don't have a fire rating?’ You get visual feedback instantly. But the real value is in what comes next:
‘For all doors without a fire rating, add FD30. Fix the syntax of the data across all objects’.
One of our customers recently used this exact workflow to update more than 15,000 doors across multiple Revit models. The completed BIM data was then synchronised back into the original Revit files using our plugin.
Instead of manually opening dozens of models or building custom scripts to edit BIM data at scale, the work was completed in minutes.
2. Geometry Quality: Add design value without draining billable hours
Whilst improving data quality will significantly increase trust, having accurate and detailed geometry enables more stakeholders to utilise the models for their workflows.
Qonic’s entire geometry kernel is available to our AI tooling, which means AI works inside the BIM model, so we can generate and adapt geometry in specific instances, or applicable to entire models.
Architects are commonly asked to model skirting board as geometry in my models. Why? Because the contractors and quantity surveyor wanted to measure the length and quantify material. Often they must push back on this request as it is both too time-consuming and negatively impacts the traditional BIM authoring tool performance. It’s time to stop simplifying models for ‘better performance’ and stop reducing model value for other stakeholders.
In Qonic, it’s simple. With our AI built in, you can ask the model: ‘Use the shadow gap profile' or 'Add a timber skirting board to all apartments’.
Kilometers (or miles if you prefer) of new geometry, accurately classified with materiality and immediately ready for use by your partners (measurement, quantity take-off, etc.). All without you having invested any time manually drawing the path room by room, saving you thousands of hours whilst increasing your model’s value.
3. Deliverables: Meeting your client requirements automatically
How much time can you spend analysing the models your company creates against your client’s requirements? This usually means jumping between authoring and review tools, building custom validation scripts for every project, and losing hours to manual audits.
Qonic automatically evaluates your model against your Client's Information Delivery Specification (IDS). Crucially, we don't just report non-compliance, we recommend solutions and, with your approval, action an instant fix across the entire model. No client IDS? You can easily create your own based on internal organizational requirements to appraise your models across each design and construction stage.
Quickly review and assess how your organisation is delivering BIM, without the significant investment of your time, opening and analysing each model manually, or with dead-end dashboards. Get better holistic oversight, identify where your models don’t meet your delivery requirements and then fix it, instantly.
If you’d like to know more about these AI use cases, or other examples of how Qonic is able to utilise automation, augmentation, generative and autonomous actions, get in touch or test Qonic today for free.
