This AI startup says it can generate permit-ready blueprints in minutes

Blueprints AI, a bootstrapped startup based in Santa Monica, California, is unveiling a second-generation version of its construction-blueprint software at TechCrunch Disrupt, which runs October 13 to 15 in San Francisco. Al Drams launched the concept roughly six months after ChatGPT's November 2022 release and officially incorporated the company in 2025 with co-founder Coco Cao. The startup was selected for the 2026 Startup Battlefield 200 cohort and will exhibit at the Disrupt event; Drams said the new version is already in use by early customers and becomes publicly available this week.
The product takes a 2D sketch, PDF or other description of a project and produces code-compliant, permit-ready architectural construction blueprints. Version 2 extends that beyond architectural drawings: from a starting description or sketch, it generates blueprints covering architecture, structural, mechanical, electrical, plumbing and interior design at once. Users can also design and make changes inside the tool, and a single design change propagates through the entire project while being checked against local building codes. That constraint-checking loop is the part that separates this kind of system from a generic image generator, because the output has to satisfy a jurisdiction's rules rather than merely look plausible. Drams said a third version is already in development, planned to add structural analysis and electrical panel design.
Pricing and packaging for the first-generation software run from $3,200 to $8,500 per month depending on the plan, with custom pricing available for enterprises. The company sells its software as a subscription and is considering a one-time flat fee for customers who need it for a single project. Blueprints AI is targeting two audiences: professional architecture, engineering and design firms on one side, and developers and real estate owners on the other.
The models behind the product were trained on eight years of approved drawing sets from Drams' own design and construction firm, which remains in operation. That is a relatively narrow but highly relevant corpus: approved sets are exactly the documents that cleared a permitting process, so they carry the formatting, annotation and code-compliance patterns that reviewers expect. A likely trade-off is geographic coverage. Building codes vary by city, county and state, and a model trained on one firm's project history will be strongest where that history is deepest; how Blueprints AI generalizes to unfamiliar jurisdictions is not something the company has detailed.
Background on the founders explains the training-data choice. Drams wrote a college thesis nearly two decades ago about software that could generate design documents for school buildings, while studying architecture in the footsteps of his parents and grandfather. He earned bachelor's and master's degrees in architecture from the University of Tehran and completed two years of Ph.D. coursework in digital design at the European Graduate School in Switzerland. He moved to the United States and opened Drams Architects in 2018, then added a construction business in 2021 after obtaining a general contractor's license. He told TechCrunch that he was not satisfied running those businesses because the impact would be limited, and that what excites him about technology is the hours and energy saved when people stop repeating the same work.
Cao, a data scientist from Hong Kong, met Drams at an AI conference several years ago. She serves as chief commercial officer, handling sales and go-to-market strategy, while Drams leads the technical and domain side.
The co-founders frame the value proposition around time allocation rather than replacement. Drams said architects spend about 80% of their time drafting and only 20% on design, and that he wants to flip that ratio to 20-80 so architects have more time to design and can produce better spaces. On the question of displacement, both co-founders said the technology does not replace licensed architects: licensed professionals still review the documents and affix their seal before the drawings are submitted for permits, and the judgment stays human. That division of labor matters in practice, because permits are issued to sealed documents and someone with a license has to stake their name on the drawings. Tools like this therefore sit in the production pipeline rather than at the signing step, at least under current rules.
This is a crowded neighborhood. Generative design and automated documentation have been an active area for AEC software vendors for years, and several recent AI startups have aimed at pieces of the workflow, from early massing studies to construction administration. What Blueprints AI is claiming is broader scope: multiple disciplines generated and kept in sync from one input. It suggests the real test will be whether the coordinated set survives professional review and local plan check without heavy manual cleanup, since that rework is where any time savings would evaporate. It is also worth noting that the headline price band, $3,200 to $8,500 per month, is a professional-software budget rather than a consumer one, so the buying decision will rest on measurable drafting hours saved per project.
Why it matters: Developers, architects and engineering firms evaluating documentation tools now have a vendor claiming multi-discipline blueprint generation with code checking, priced as a monthly subscription in the low thousands to mid-eight thousands of dollars. If the second-generation output holds up under licensed review, the practical change for teams is compressing drafting hours and shifting staff time toward design and review; if plan-check corrections stay frequent, the savings shrink. Product teams building in adjacent AEC software should also read the founders' framing as a positioning signal: augmentation of licensed professionals, not automation of the seal.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
Blueprints AI, a bootstrapped startup based in Santa Monica, developed software that produces code-compliant construction blueprints from a photo, PDF, or sketch.