
Blueprints AI, a Santa Monica startup co-founded by architect Al Dram and data scientist Coco Cao, is presenting its second-generation drafting software at TechCrunch Disrupt 2026. The event runs October 13-15 in San Francisco, where real estate startups will present new tools and services.
Dram began exploring software that could make school design documents for his college thesis, nearly two decades ago. ChatGPT launched in November 2022, and he launched the Blueprints AI concept within six months. Dram estimates architects spend about 80% of their time drafting and 20% designing. His stated aim is to give architects more time for design by reducing repetitive drafting work.
Key Takeaways
- Version 2 speeds up construction-document drafting.
- Firms tailor documents with their stored standards and past work.
- Generation times vary with project size and scope.
- Licensed professionals and building departments retain key review roles.
- TechCrunch Disrupt 2026 runs October 13-15 in San Francisco.
Select This AI Startup Says It Can Generate Permit-Ready Blueprints in Minutes
A firm can begin with a sketch, site photo, PDF, point-cloud file, or plain-text project description, without converting the file first. That flexibility lets teams start with materials they already have, rather than building a new file set first.
After the firm configures its Knowledge Base, project setup takes minutes. The Knowledge Base is a firm’s stored standards and past projects that Blueprints AI uses to tailor its generated documents. That setup comes before each project begins.
What teams can upload
The software accepts five types of project material, including early design notes and digital site information. Teams can use one input or combine several as the project develops.
- Sketches
- Site photos
- PDFs
- Point-cloud data
- Plain-text descriptions
Point-cloud data captures a site as a set of measured points. This can help teams work from a digital record of existing conditions.
How a firm’s standards shape a project
Blueprints AI takes about two weeks to onboard a firm. During that period, the model is tailored to past projects, project types, local jurisdiction comments, and company standards.
Once that work is done, users can start a new project in minutes. The configured Knowledge Base helps shape the generated design documents around the firm’s own way of working.
The company’s generation-time claims come from internal tests on a small single-family project. Those results do not set a timing promise for larger or more complex work. Treat any minutes-based generation estimate as a starting point, since project size and scope affect timing.
A firm must configure its Knowledge Base before it can set up a project in minutes.
Accepted input types include both early sketches and plain-text descriptions, so teams can begin before every drawing is complete.
What Does Version 2 Add to the Drafting Process?
Version 2 expands the output from architectural documents to coordinated documents spanning six building disciplines, including interior design. The wider scope helps teams create a broader set of construction documents from one project.
Blueprints AI trained its models on approved drawing sets from Dram’s design and construction firm, covering eight years of work. Version 2 produces architecture, structural, mechanical, electrical, plumbing, and interior documents.

One coordinated set across disciplines
Version 1 produced architectural construction documents from a sketch or PDF. The newer version starts from a project description or sketch, then extends the output to other building systems.
The comparison shows how the workflow has changed:
| Attribute | Version 1 | Version 2 |
|---|---|---|
| Starting point | 2D sketch or PDF | Project description or sketch |
| Document scope | Architectural construction documents | Architecture, structural, mechanical, electrical, plumbing, and interior documents |
| Design changes | Regenerate documents from scratch | Change updates the project documents |
Changes that carry through the project
In Version 1, a design change meant generating documents again from scratch. Version 2 updates the project documents after a change while aiming to preserve the design intent.
Code checks accompany those updates and cite applied provisions. They cover IBC, CBC, IFC, NFPA, ADA, local zoning rules, fire needs, and access needs. A code check does not ensure that a building department will approve a set.
Title 24 energy compliance, fire and life-safety documents, and civil and landscape generation come with Enterprise plans. These added options extend the document set for firms with broader project needs.
Structural analysis and electrical-panel design are planned additions for Version 3, not capabilities to describe as already released.
What Must Firms Check Before Submitting AI-Generated Plans?
Before submitting AI-generated plans, confirm stamp and plan-acceptance rules with the building department reviewing them. The answer depends on the project and jurisdiction, since a permit set may need review and a seal from a state-licensed architect or engineer.
The local building department is the authority on what it will accept. It can confirm these four points:
- Whether the set needs a professional stamp.
- Whether size or complexity thresholds apply.
- Which licensed professional must review and seal the set.
- Whether the department accepts plans produced elsewhere.
“Permit-ready” describes a document workflow, not a promise that the building department will approve a set.
Blueprints AI’s Version 1 subscription costs between $3,200 and $8,500 monthly, depending on the plan, with custom Enterprise pricing. That price does not remove local review steps or the need to check local building codes.
FloorPlan-AI serves a narrower purpose: it generates editable floor plans and exports DWG and DXF files. Its options include four free credits, a one-time Project Pack with 30 credits for $9.9, and a $25 monthly Pro plan with 100 credits. It does not offer the same coordinated construction-document workflow across disciplines.
Keep the licensed professional’s review and the building department’s requirements in the submission workflow.
Conclusion
Blueprints AI’s value is faster document production, while project-specific review remains part of preparing a permit set for submission. The software can help firms move from design material to construction documents, but it does not decide whether a local building department will accept them.
Firms should compare the software with their own standards, project types, and design needs. Before preparing a permit submission, they should contact their local building department and licensed design professional.
Firms considering the platform should first map their document needs and confirm local submission requirements with their licensed professional.



