AI startup Blueprints AI says it can turn sketches into permit-ready plans in minutes

Blueprints AI, a young architecture software startup founded by architect Al Drams and data scientist Coco Cao, says it can turn a 2D sketch or PDF into code-compliant, permit-ready blueprints in minutes. The company will publicly showcase its upgraded platform this week at TechCrunch Disrupt 2026 in San Francisco, where it hopes to demonstrate how generative AI can remove much of the repetitive drafting work from architecture.

The timing matters because the construction and design industries still depend on time-consuming document production, multiple revisions and code checks before a project can move forward. Blueprints AI says its software can speed up that process dramatically while leaving final sign-off to licensed professionals.

What Blueprints AI says it does

Blueprints AI is building software that generates architectural construction documents from a simple starting point, such as a hand sketch, a PDF or a written project description. The company says its system can produce blueprints that are not only detailed but also aligned with local building codes and suitable for permit review.

The startup’s pitch is straightforward: what once took days or weeks of drafting work could be condensed into a process measured in minutes. Drams and Cao say that kind of speed-up could change how architects, engineers, developers and property owners approach early-stage planning.

Why this matters for architects

Drams argues that architecture professionals spend most of their time on repetitive drafting rather than higher-value design work. In his view, AI should not replace the creative and licensed judgment of architects; it should reduce the amount of manual production work that slows them down.

The company says its goal is to shift the balance of an architect’s day away from paperwork and toward actual design thinking. That, it argues, could improve both productivity and the quality of the built environment.

Drams says architects spend roughly 80% of their time drafting and only 20% designing, and he wants to invert that ratio so professionals can devote more energy to design.

How an old thesis idea became a startup

Drams’ path to Blueprints AI stretches back nearly two decades. As a university student studying architecture, he wrote a thesis on software that could generate design documents for school buildings. The idea was ambitious, but the underlying technology needed to make it practical did not yet exist.

That changed, he says, when ChatGPT and other large language models arrived. The generative AI wave gave him a new way to revisit an old concept and turn it into a commercial product.

Within about six months of seeing what modern AI tools could do, Drams launched the idea that became Blueprints AI. The company was formally incorporated in Santa Monica in 2025 with Cao as co-founder.

The founders’ backgrounds

Drams studied architecture in Iran, earning both a bachelor’s degree and a master’s degree from the University of Tehran. He later completed two years of doctoral coursework in digital design in Switzerland before relocating to the United States.

He eventually opened Drams Architects in 2018 and later obtained a general contractor’s license, which allowed him to launch a construction business in 2021. That dual experience in design and construction helped shape the startup’s approach to software development.

Cao, who studied and worked in data science, met Drams at an AI conference several years ago. She now serves as chief commercial officer and leads sales and go-to-market strategy.

How Blueprints AI trained its system

The company says its models were trained using eight years of approved drawing sets from Drams’ architecture and construction firm, which remains active. Those documents gave the startup real-world examples of plans that had already passed through review and approval processes.

That kind of training data is important because architectural software has to deal with much more than visual layout. It must account for code constraints, technical systems, spatial relationships and coordination across disciplines.

Blueprints AI is positioning itself as a workflow tool rather than a replacement for licensed human expertise. According to the founders, final drawings still need to be reviewed, stamped and filed by professionals before they can be submitted for permits.

What is new in version 2?

The startup says the second generation of its product is a significant upgrade over the first. It will allow users to begin with a sketch or written description and generate a full set of project plans covering architecture, structure, mechanical systems, electrical work, plumbing and interior design.

Another key change is editability. Users can now modify a design inside the system, and when one part of the plan changes, the update propagates through the rest of the project. The software then checks the revised design against local codes.

That matters because building projects are rarely static. A change in one area can affect everything else, from load-bearing elements to system layouts and interior constraints. Blueprints AI says its software is designed to keep the whole package consistent as revisions happen.

Who is Blueprints AI targeting?

The company says it is serving two main customer groups. The first is professional firms in architecture, engineering and design. The second is developers and real estate owners who need planning documents as part of a project pipeline.

That split reflects two sides of the market: firms that produce construction documentation as part of their core work, and property stakeholders who want to move projects along faster. Blueprints AI believes both audiences can benefit from shorter turnaround times and more consistent coordination.

How the business model works

Blueprints AI currently sells access through a subscription model. For organizations with ongoing project demand, monthly pricing may make the software easier to budget for and integrate into standard workflows.

The startup is also considering a one-time flat-fee option for customers who only need the tool for a single project. That could broaden adoption among smaller developers or owners with limited one-off use cases.

Its existing product has been priced at between $3,200 and $8,500 per month depending on the plan, with custom enterprise pricing also available.

Why the second version is a bigger deal than the first

Version 2 is important because it moves Blueprints AI beyond simple document generation into a more complete design workflow. Instead of producing static output, the software now supports edits, cross-discipline coordination and code checks after changes are made.

In practical terms, that means an architect or designer could start with a concept and move through more of the documentation process inside one platform. The company hopes that makes the software useful not just as a drafting shortcut, but as an operational layer in the design process.

Drams and Cao also say the second generation is already being used by early customers and will be open to the public this week. That public rollout gives the company a chance to prove it can serve real projects at scale, not just pilot users.

What comes next for Blueprints AI?

The startup is already building a third version of its platform. According to the founders, that future release is expected to add structural analysis and electrical panel design, expanding the software’s role deeper into engineering-related tasks.

If that roadmap holds, Blueprints AI could evolve from a drafting assistant into a broader design and analysis platform. That would put it closer to the center of pre-construction workflows, where speed, precision and coordination are critical.

Even so, the company is careful not to frame its product as an autonomous replacement for licensed experts. Instead, it presents itself as a tool that compresses the amount of manual work needed before those experts step in.

How does it fit into the broader AI wave?

Blueprints AI is part of a much larger trend: vertical AI startups that use large language models and related systems to automate specialist workflows in tightly defined industries. Rather than building a general-purpose chatbot, the company is targeting a high-value professional niche with clear document standards and measurable time savings.

That focus may be one reason the startup made it into TechCrunch’s Startup Battlefield 200, a showcase for early-stage companies with breakout potential. Disrupt, which runs from October 13 to 15 in San Francisco, is designed to put those startups in front of investors, customers and the wider tech industry.

For architecture, the opportunity is obvious. Design firms work with detailed regulations, complex revisions and expensive labor, all of which create a strong incentive for tools that can automate the routine parts without removing professional oversight.

Why the human review still matters

Blueprints AI’s founders say licensed architects remain central to the process. Even if software drafts the documents, a professional still needs to review them, make sure they are appropriate for the project and submit the final package under seal.

That distinction is important for both legal and practical reasons. Building codes vary by location, projects can be unusually complex, and final responsibility still rests with humans who are licensed to do the work.

In other words, the software may accelerate production, but it does not eliminate accountability.

Timeline: from thesis to Disrupt demo

1. Early 2000s: Drams develops a thesis concept for software that could generate design documents for school buildings.
2. 2018: He opens Drams Architects in the United States.
3. 2021: He adds a construction business after becoming a licensed general contractor.
4. November 2022: ChatGPT’s rise convinces him the original idea may now be feasible.
5. 2025: Blueprints AI is incorporated in Santa Monica with Coco Cao as co-founder.
6. 2026: The startup is selected for Startup Battlefield 200 and prepares to launch version 2 at Disrupt in San Francisco.

What investors and customers will watch for

The most important question for Blueprints AI is whether the product can deliver reliable, code-aware output across different project types and jurisdictions. Generating a polished blueprint is one thing; generating a blueprint that survives real-world review is something more demanding.

Potential customers will likely look at three things: accuracy, ease of revision and how much human oversight is still required. If the software can significantly reduce drafting hours without creating more review work, it could prove compelling.

Investors, meanwhile, will probably want to know whether the startup can scale beyond its founder-led data and whether it can expand into adjacent workflows without losing precision. Architecture software can be sticky if it becomes part of a firm’s day-to-day process, but only if it consistently saves time.

The bigger picture

Blueprints AI is not just another generative AI experiment. It is a focused attempt to modernize one of the more document-heavy corners of the built-environment economy. If the company’s claims hold up, it could help shift architecture work from repetitive drafting toward higher-level design decisions.

That is an appealing promise in a profession where projects are often slowed by manual coordination, code interpretation and document churn. Whether the startup can deliver that promise at scale will become clearer as version 2 reaches more users this week.

For now, the company’s message is simple: AI should not replace architects, but it may finally give them back the time to do more of the work that matters most.