ArchiLabs learning guides

Put AI to work in your CAD and BIM workflows.

Choose the work you need to do. These guides cover the inputs, modeling steps, and review decisions behind useful automation in Revit, AutoCAD, and connected project tools.

Explore ArchiLabs Desktop Browse all articles

Get started with AI in Revit and AutoCAD

Start with a task you already understand: a parameter update, drawing correction, or model query. Give the assistant the source information and define what a successful result looks like. A small, repeatable workflow makes it easier to compare the effort of doing the work and reviewing it.

Connect Revit parameters, schedules, and spreadsheets

Model data work starts with matching the right record to the right element. Establish the identifier, field mapping, units, and approved source before a batch update. Keep unmatched rows and conflicting values visible so the next person can resolve them.

Turn drawings and specifications into useful BIM

A drawing or equipment PDF can supply valuable modeling context, but it does not answer every question about scale, levels, families, or parameters. Work from explicit references, record assumptions, and inspect the native elements as well as the finished view.

Automate AutoCAD drafting and drawing updates

Separate editable drawing objects from references and images before assigning a CAD task. For redlines, layers, attributes, and title blocks, describe the intended change and preserve a reference to the original drawing. Check the plotted result along with the underlying objects.

Carry project information across connected tools

Files, meeting decisions, and model data often describe different stages of a project. Choose which source governs each change, separate decisions from suggestions, and give reviewers a concise account of completed work and unresolved questions.

Apply AI to equipment, routing, and engineering documentation

Use a specific equipment or routing task to connect design intent with modeling work. Define interfaces, constraints, and deliverables, then have the responsible team review the result. Model completion and engineering acceptance are separate checkpoints.