Drafted

Cover Block

Open sources

Field	Value
Name	Drafted
Tagline	AI-powered platform generating residential floor plans, exterior designs, and 3D layouts from natural language.
Founded	2026
Stage	Seed
Business model	B2C
Industry	Proptech
Technology	AI / Machine Learning
Growth profile	Venture Scale
Founding team	Repeat Founder
Funding label	Seed
Total disclosed funding	~$17,650,000 [TechCrunch, December 2025]; [Business Insider, May 2026]

Links

Open sources

• Website: https://www.drafted.ai/

What an Investor Needs First

Open sources Drafted is building an AI home-design platform that turns natural-language prompts, sketches, and spatial constraints into residential floor plans, exterior concepts, and 3D layouts, and it merits attention now because it has already attracted a sizeable seed round and early usage that is notable for a company founded in 2026 [Business Insider, May 2026] [Y Combinator] [Drafted]. The founding story is unusually direct: founder and CEO Nick Donahue is revisiting the same housing-design problem after previously co-founding Atmos, a custom-home-design startup that TechCrunch profiled in 2020 and that later became the backdrop for Drafted’s second attempt at lowering the cost and friction of custom home design [TechCrunch, 2020-06-25] [TechCrunch, December 2025].

The product appears designed to sit between consumer inspiration tools and professional design workflows, with Drafted describing plan generation, comparison, 3D rendering, and exports into CAD, BIM, DXF, IFC, and Revit-compatible workflows, which suggests the differentiation is not only image generation but structured layout output that can move closer to downstream design software [Drafted] [Drafted] [Y Combinator]. That matters because many AI design tools stop at concept imagery, while Drafted is at least positioning itself around editable plans and file interoperability, a more demanding product surface if it works as described [Drafted] [Y Combinator].

On the team side, the main public datapoint is Donahue’s prior category exposure rather than a fully disclosed bench, and that cuts both ways: investors can point to domain persistence and prior operating experience, but the public record still says more about the founder than about repeatable commercial execution inside this new entity [TechCrunch, December 2025] [TechCrunch, 2020-06-25]. Funding is clearer than operations: TechCrunch reported a $1.65 million pre-seed in December 2025, and Business Insider reported a $16 million seed in May 2026 led by Buckley Ventures, with investors including Y Combinator, Ben Silbermann, and Ryan Tedder, putting disclosed funding at roughly $17.65 million for what is described as a B2C product today [TechCrunch, December 2025] [Business Insider, May 2026].

The near-term question is less whether there is consumer curiosity, the company and Business Insider both indicate there is, and more whether Drafted can convert plan generation into durable paid behavior, professional adoption, or transaction adjacency without losing speed or usability [Business Insider, May 2026] [Drafted]. Over the next 12 to 18 months, the public markers worth watching are whether user activity remains strong beyond launch novelty, whether the product earns trust from architects, builders, or developers rather than only homebuyers, and whether its current free-to-use posture evolves into a credible monetization model [Connected Design, 2026] [Business Insider, May 2026] [Drafted].

Claim stands unchecked -- This section relies on a mix of named-publisher reporting and company-published product and usage claims, with several material operating metrics uncorroborated outside company-linked sources.

Taxonomy Snapshot

Axis	Value
Stage	Seed
Business Model	B2C
Industry / Vertical	Proptech
Technology Type	AI / Machine Learning
Growth Profile	Venture Scale
Founding Team	Repeat Founder
Funding	Seed, total disclosed approximately $17.65 million

Inside the Company

PUBLIC

Drafted presents itself as an AI home-design company focused on residential layouts rather than general architectural software. On its website, the company says users can create floor plans, compare layouts, explore 3D renders, collaborate on projects, and export CAD and BIM files, which is enough to place the product squarely in consumer-adjacent proptech with some workflow ambitions into professional design tools [Drafted]. Crunchbase describes the company in similar terms, as a platform for personalized home floor plans and layouts based on user inputs [Crunchbase].

The founding record is still thin in public source material, but the available company profile points to a 2026 founding date and identifies Nick Donahue as founder [Crunchbase]. Y Combinator lists Drafted in its company directory and describes the product as a tool that lets users design homes by combining sketches and direct spatial input to generate floor plans and elevations [Y Combinator]. Headquarters and legal entity details were not surfaced in the cited company and database records used here.

Chronology matters here because the public footprint formed quickly. Crunchbase lists the company as a seed-stage business [Crunchbase], while the website already presents a live product surface with usage counters and export features [Drafted]. Taken together, the public record suggests a company that moved from formation to broad public product availability on an unusually compressed timeline, though the underlying operating milestones remain only partially documented in company-controlled sources [Drafted] [Crunchbase].

Partially corroborated -- Relies primarily on company website, Crunchbase, and Y Combinator directory information, with limited independent public corroboration on founding details.

Under the Hood

MIXED The product story is clearer than most seed-stage proptech software because the company has shown the workflow in public, even if the underlying model stack remains mostly undescribed. Drafted presents itself as an AI home-design tool that generates residential floor plans, 3D models, renders, and CAD or BIM exports from user inputs, while Y Combinator describes the core interaction more specifically as combining sketches and direct spatial input to generate floor plans and elevations [Drafted] [Y Combinator]. The public-facing product also supports layout comparison, project collaboration, and file export in formats including PDF, AutoCAD DXF, and IFC, with a Revit-based generation flow referenced on the company site [Drafted].

The user workflow appears oriented around constrained generation rather than open-ended image synthesis, which matters for whether the product can fit real design work. According to Drafted's own product pages, users can specify bedrooms, bathrooms, living spaces, house shape, and target size, then refine generated plans by adjusting walls, doors, and windows or regenerating selected areas [Drafted]. The site also shows paired outputs, a schematic plan and a furnished view, plus exterior renders with selectable materials, colors, and architectural styles, suggesting the product is trying to span early ideation and presentation rather than stopping at a single floor-plan draft [Drafted].

What remains less verified is the technical architecture beneath that experience. Public materials support that the system accepts structured constraints such as room lists, lot requirements, square footage targets, sketched footprints, and optional room placements, but they do not disclose the model providers, training data, or how reliably generated plans map to buildable code-compliant documentation [Drafted] [How to Use an AI Floor Plan Generator | Drafted]. For investors, that leaves the product differentiated at the workflow and export layer in public evidence, while the defensibility of the model layer still depends largely on company assertions and future proof points in professional adoption [Business Insider, May 2026] [TechCrunch, December 2025].

Claim stands unchecked -- This section relies materially on Drafted's own website and product pages, with partial corroboration from Y Combinator and named-press descriptions.

Market Research

PUBLIC

The market matters now because generative design has arrived just as housing affordability, permitting friction, and consumer appetite for faster pre-construction decision-making are all pushing more of the home-design workflow into software, even if Drafted itself does not yet sit in a neatly reported market category [Business Insider, May 2026] [TechCrunch, December 2025].

The first constraint here is evidentiary. No named third-party market report in the supplied research provides a direct TAM, SAM, or SOM for AI-native residential floor-plan generation, so any sizing exercise has to lean on adjacent categories rather than present Drafted's addressable market as measured fact. The closest public analogs in the source set are custom-home design and broader digital home-planning software, inferred from coverage describing Drafted as a bid to make custom-home design faster and cheaper and as a tool for generating plans, renders, and CAD or BIM outputs [TechCrunch, December 2025] [Drafted]. That framing suggests Drafted is not competing only for architecture software spend. It is also competing for consumer intent earlier in the funnel, before a buyer hires an architect, selects a stock plan, or commits to a builder [Business Insider, May 2026].

Demand appears to be pulled from that upstream consumer moment. Business Insider reported 250,000 website visitors since launch and more than 300,000 floor plans generated in one recent month, while also noting that more than one-third of visitors were homebuyers [Business Insider, May 2026]. Those figures are company-reported usage metrics rather than audited commercial results, but they do help clarify where early interest may be concentrated: not just among professional designers, but among consumers using software to test feasibility, layout preferences, and style choices before entering a higher-cost design process [Business Insider, May 2026].

A second tailwind is workflow compression. Drafted's own product materials emphasize natural-language inputs, sketch-based plan generation, editing tools, and exports into PDF, DXF, IFC, and Revit-linked workflows, which places the product near the boundary between consumer ideation and professional execution [Drafted] [Y Combinator]. If that handoff proves reliable, the adjacent markets extend beyond home-design inspiration tools into parts of the architecture, engineering, and construction software stack, especially early schematic design and client collaboration. The substitute set is broad: traditional architects, stock house-plan libraries, conventional CAD or BIM software, and open-source or desktop design tools that serve cost-sensitive users with more manual control [Drafted] [Flowtivity].

Macro conditions cut both ways. TechCrunch's December 2025 coverage framed Drafted as founder Nick Donahue's return to the same problem after Atmos, a prior housing startup that struggled as higher interest rates and weaker client demand weighed on the custom-home market [TechCrunch, December 2025]. That history matters because residential construction software demand is still exposed to the housing cycle, financing costs, and consumer confidence. At the same time, the same pressure can support software adoption if homeowners, builders, and designers are pushed to reduce iteration time and pre-construction costs. Regulatory forces also matter indirectly. Any product that influences floor plans and buildable layouts still feeds into a permit, code, and licensed-design environment, which means software enthusiasm does not remove downstream compliance constraints, even where it may speed concept generation.

Market lens	What the sources support	Evidence quality
Direct market size	No named third-party TAM, SAM, or SOM for AI residential floor-plan generation appears in the supplied research	Low [Business Insider, May 2026] [TechCrunch, December 2025]
Adjacent category	Custom-home design and digital home-planning workflows are the closest supported analogs	Medium [TechCrunch, December 2025] [Drafted]
Early demand signal	Consumer-led interest appears meaningful, with more than one-third of visitors reportedly homebuyers	Low to medium, company metric cited by publisher [Business Insider, May 2026]
Substitute markets	Traditional architects, stock plans, CAD or BIM tools, and open-source design software remain viable alternatives	Medium [Drafted] [Flowtivity]

The practical reading is that the market case rests less on a published category map and more on whether Drafted can turn consumer design intent into a repeatable software workflow that professionals will trust. That is a plausible opening, but the public evidence still describes interest and workflow adjacency more clearly than it describes market size with precision.

Partially corroborated -- Section relies on named-publisher reporting from Business Insider and TechCrunch, plus company and YC materials for workflow scope. No independent third-party market-sizing report was available in the supplied sources.

Competition and Substitutes

MIXED Drafted sits between consumer floor-plan generators and professional design software: it is trying to make early-stage residential design easier for homebuyers while still exporting into architect and builder workflows [Business Insider, May 2026] [Drafted] [Y Combinator].

Company	Positioning	Stage / Funding	Notable Differentiator	Source
Drafted	AI platform for residential floor plans, exterior designs, and 3D layouts from prompts, sketches, and spatial inputs	Seed, $17.65M disclosed across pre-seed and seed rounds	Bridges consumer-style generation with CAD, BIM, IFC, DXF, and Revit workflow hooks	[Business Insider, May 2026] [TechCrunch, December 2025] [Drafted]
HouseGAN++	Research model for graph-constrained floor plan generation	Research project	Native focus on generating floor plans from relational constraints rather than a packaged end-user home-design product	[Flowtivity, retrieved 2026]
Graph2Plan	Research system for floor plan generation from graph representations	Research project	Structured graph-based approach to room layout generation	[Flowtivity, retrieved 2026]
Sweet Home 3D	End-user home interior and layout design software	Established software product	Conventional design workflow with a broad installed base rather than prompt-led generation	[Flowtivity, retrieved 2026]
IfcOpenShell	Open-source toolkit around IFC workflows	Open-source project	Deep relevance to interoperable building-model workflows rather than consumer-first ideation	[Flowtivity, retrieved 2026]

The field breaks into three visible segments. At one end are research-led generation systems such as HouseGAN++ and Graph2Plan, which matter less as direct commercial rivals than as evidence that layout generation itself is becoming easier to reproduce [Flowtivity, retrieved 2026]. At the other end are established or open tools such as Sweet Home 3D and IfcOpenShell, which serve users who are willing to trade convenience for control, lower cost, or workflow interoperability [Flowtivity, retrieved 2026]. Drafted is aiming for the middle: a prompt-driven front end that feels accessible to non-experts, with export paths into professional formats once a concept is worth refining [Drafted] [Y Combinator].

Drafted's current edge appears to be product packaging and founder context, not a clearly verified proprietary moat. The product surface described on its site is broader than a simple image generator, spanning floor plans, furnished views, exterior renders, collaborative projects, and exports to PDF, DXF, IFC, and Revit-linked workflows [Drafted]. That makes the wedge more credible for actual residential planning use, while Nick Donahue's prior work at Atmos gives the company a founder who has already spent years on the same custom-home-design problem set [TechCrunch, 2020-06-25] [TechCrunch, December 2025]. The issue is durability: interfaces and workflow integrations can matter in the near term, but absent evidence of exclusive data, exclusive channel access, or embedded enterprise adoption, that edge is perishable if adjacent design tools add similar AI generation features.

The main exposure is that Drafted does not obviously control the professional channel yet, while open and research alternatives compress technical differentiation. Sweet Home 3D and similar traditional tools benefit from familiarity and a more established user expectation around editing and precision, even if they lack Drafted's AI-first workflow [Flowtivity, retrieved 2026]. IfcOpenShell points to another limit: interoperability and standards-based building-model work are their own domain, and being able to export IFC files is not the same as owning that ecosystem [Drafted] [Flowtivity, retrieved 2026]. Drafted is also exposed if architects and builders decide that prompt-based ideation belongs upstream of, not inside, their core production stack.

The most plausible 18-month scenario is a sorting between inspiration-layer products and production-layer tools. Drafted is the likely winner if consumer and prosumer users keep treating home design as a self-serve entry point, because its mix of natural-language generation and downstream exports is better aligned to that workflow than HouseGAN++ or Graph2Plan, which are better understood as technical approaches than scaled products [Business Insider, May 2026] [Flowtivity, retrieved 2026]. The likely loser if interoperability and standards compliance become the buying center is Drafted rather than IfcOpenShell, because open infrastructure tends to strengthen when the market values portability over convenience [Drafted] [Flowtivity, retrieved 2026]. That leaves the company in an interesting but still early position: well placed to capture ideation demand, not yet shown to own the professional system of record.

Partially corroborated -- This section relies on named public reporting from Business Insider and TechCrunch for Drafted, plus company materials for product claims. Competitor detail is thinner and rests primarily on a third-party comparative source, so the section is only partially corroborated.

Opportunity

Upside path

PUBLIC The prize here is not a better floor-plan toy, but a software layer that could sit at the front of residential design and capture intent before architects, builders, or homebuyers commit to a plan.

The headline opportunity is straightforward. If Drafted can turn free-form prompts, sketches, and spatial constraints into usable floor plans, exterior concepts, 3D renders, and CAD or BIM exports, it has a chance to become the default starting point for residential design work, especially for custom homes and early-stage concepting [Y Combinator] [Drafted]. That outcome is still early, but it is not purely aspirational: TechCrunch reported a $1.65 million pre-seed in December 2025, and Business Insider reported a $16 million seed in May 2026 led by Buckley Ventures, with Y Combinator and named angel backers also involved [TechCrunch, December 2025] [Business Insider, May 2026]. The same Business Insider report said the company had reached 250,000 website visitors since launch and that users generated more than 300,000 floor plans in one month, which suggests there is already meaningful top-of-funnel demand for faster home-design workflows, even if monetization and retention remain unproven [Business Insider, May 2026].

The clearest growth paths are visible in the product surface that is already public: Drafted is not limited to image generation, but extends into editing, collaboration, and export formats that can connect concept generation to professional workflows [Drafted] [Y Combinator]. That gives it more than one route to scale.

Scenario	What happens	Catalyst	Why it's plausible
Consumer design entry point	Drafted becomes the default place for homebuyers and aspiring custom-home customers to generate, compare, and iterate on layouts before they hire an architect or builder.	Sustained consumer usage around floor-plan generation and renders, building on the reported 250,000 visitors since launch and 300,000-plus plans generated in a month [Business Insider, May 2026].	The product already lets users specify room counts, size targets, shape, materials, and style, then compare layouts and explore 3D renders, which fits the earliest phase of residential purchase intent [Drafted] [Jawlah].
Pro workflow bridge	Drafted wins architects, builders, and developers as a concepting and pre-production layer because outputs can be edited and exported into standard design tools.	Adoption of CAD, DXF, IFC, and Revit-connected workflows, which would move the product from inspiration into production-adjacent use cases [Drafted].	The public product claims include project collaboration, wall and window editing, and export to PDF, AutoCAD DXF, IFC, or direct generation inside Revit, all of which make professional adoption at least technically reachable [Drafted].
Category platform in residential design data	Drafted compounds usage into a proprietary dataset on how consumers and professionals describe, modify, and select home layouts, improving generation quality and recommendation loops over time.	More iterative usage and larger design-volume accumulation, supported by the company-reported base of 148,630 users and 11,322,730 plans generated [Drafted].	If those company-reported usage figures are directionally accurate, the company may already be collecting a large body of structured design interactions, which is the kind of asset that can improve model performance and user relevance over time [Drafted].

What compounding looks like here is a product-data-distribution loop. A user starts with a prompt or sketch, edits walls and openings, compares layouts, renders the design, shares it with collaborators, and exports it to downstream software, each step generating feedback about what was requested, what was changed, and what ultimately looked usable [Drafted] [How to Use an AI Floor Plan Generator | Drafted]. If Drafted can learn from those iterations, then every design session can improve default layouts, room adjacencies, style presets, and export readiness. That matters because residential design has a high-friction starting point, and a tool that gets meaningfully better with volume can become hard to displace.

There is also a founder-specific reason the upside is more reachable than a cold-start idea. TechCrunch's 2025 profile framed Drafted as Nick Donahue's second attempt at affordable custom-home design after Atmos, and Business Insider tied the new company to the same underlying problem set [TechCrunch, December 2025] [Business Insider, May 2026]. Second-time founders do not guarantee execution, but prior category exposure can shorten product intuition, hiring, and go-to-market iteration, particularly in a workflow-heavy category where user trust depends on details rather than broad consumer awareness.

The size of the win is best framed conditionally, because no confirmed public market sizing was provided in the source set. A reasonable comparable on outcome shape, not valuation parity, is Autodesk's long-standing position as workflow infrastructure for design professionals, while a more immediate analog is any software company that owns the handoff between ideation and production in a high-value offline spend category [Drafted]. If Drafted were to become a standard entry point for residential design intent and the handoff layer into architectural or builder workflows, it could plausibly support a multibillion-dollar enterprise value over time (scenario, not a forecast), because it would be monetizing the software control point that sits upstream of a very large construction and renovation spending base. The public evidence does not yet support precision beyond that. What it does support is a narrower claim: Drafted has already shown enough financing support, product breadth, and early usage to make a platform outcome conceivable rather than theoretical [TechCrunch, December 2025] [Business Insider, May 2026] [Y Combinator] [Drafted].

Claim stands unchecked -- This section relies on named-publisher funding coverage from Business Insider and TechCrunch, but several material traction and product-depth claims come from company-controlled sources and Y Combinator listings rather than independently corroborated operating data.

Sources

Open sources

1. [TechCrunch, December 2025] This founder just landed backing for a second go at the same problem: affordable custom home design | https://techcrunch.com/2025/12/23/this-founder-just-landed-backing-for-a-second-go-at-the-same-problem-affordable-custom-home-design/

2. [Business Insider, May 2026] An AI startup raised $16 million to make designing a home feel more like vibe coding | https://www.businessinsider.com/drafted-16m-funding-reshape-home-design-with-ai-2026-5

3. [Y Combinator] Drafted: Design your home instantly with AI | https://www.ycombinator.com/companies/drafted

4. [Drafted] AI House Plan Generator | Drafted | https://www.drafted.ai/

5. [TechCrunch, 2020-06-25] Atmos wants to make building a house a one-click effort | https://techcrunch.com/2020/06/25/atmos-wants-to-make-building-a-house-a-one-click-effort/

6. [Connected Design, 2026] Putting AI to Use in Floor Plan Generation with Drafted | https://connecteddesign.tech/putting-ai-to-use-in-floor-plan-generation-with-drafted/

7. [Crunchbase] Drafted - Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/drafted-b9b3

8. [Flowtivity, retrieved 2026] Drafted.ai Review 2026: Free AI House Plans vs the Open Source Alternatives | https://flowtivity.ai/blog/drafted-ai-open-source-alternatives/

9. [How to Use an AI Floor Plan Generator | Drafted] How to Use an AI Floor Plan Generator | Drafted | https://www.drafted.ai/learn/ai-floor-plan-generators

10. [Jawlah] Drafted raises $16 million for AI-powered home design | https://jawlah.co/en/57814