Davis is an AI-native real estate company that delivers feasibility studies and architectural designs in days rather than months. The company combines proprietary generative AI models with in-house architect expertise to produce regulatory-compliant floor plans, financial modeling, and site analysis as a full-stack service. Its Gaudi-1 model generates structured architectural layouts using discrete diffusion, achieving state-of-the-art results on industry benchmarks.
Funding
Funding not disclosed
Founders
Product
Problem
Real estate development is one of the world's largest industries, yet early-stage workflows remain fragmented and slow, often taking weeks or months to complete. Projects require coordination across multiple firms and stakeholders, with feasibility studies and architectural designs delivered as disconnected, sequential deliverables. This inefficiency carries direct financial costs, particularly for debt-funded projects where time has a measurable price.
Solution
Davis delivers real estate development work as a service, combining proprietary AI systems with human architectural judgment. The company produces feasibility studies covering site analysis, regulatory diagnostics, volumetrics, and financial modeling, alongside architectural designs including floor plans, unit mix optimization, and space planning. Its Gaudi-1 model uses a discrete diffusion approach to generate buildings as structured compositions of architectural elements, with regulatory constraints integrated directly into the generation process rather than applied as post-processing. In-house architects refine and validate every output before delivery, compressing what traditionally required multiple firms and weeks of coordination into a single integrated process completed in days.
Target Audience
Primary customers are real estate developers, property investors, and development firms seeking faster feasibility assessments and architectural design outputs without managing multiple external consultant engagements.
Features
- Gaudi-1 proprietary generative model using discrete diffusion for structured architectural layout generation
- Constraint-aware generation that integrates regulatory, site, and market requirements directly into the design process
- State-of-the-art performance on RPLAN and MSD benchmarks across IoU, FID, and KID metrics
- Parallel analysis of project scope across regulations, market data, site constraints, and architectural inputs
- Feasibility study delivery including site analysis, regulatory diagnostic, volumetrics, and financial modeling
- Architectural design outputs including floor plans, unit mix optimization, and space planning, with in-house architect validation