AIRI lab

AIRI lab is an AI image and design platform trained for the built environment. Unlike general image generators, its models are tuned on architectural project data annotated by architects, so outputs follow spatial composition, site logic and material language. It runs in the browser and covers work from early massing studies to presentation images and short videos.

Key features

• Smart design assistant – an AI agent that reads a brief, extracts key concepts and styles and suggests creative directions
• AI image creation – generate concept visuals and polished renders from sketches, text prompts or references
• Inspire, inpaint and upscale – edit parts of an image, combine styles and output high-resolution results
• Templates – scene templates for massing, facade and site design, interior layouts and materials, parks and zoning or mobility studies
• Variations – turn one design into ten or more material, lighting or mood options
• Still to motion – convert static designs into video and shareable presentations
• Team features – shared asset libraries, access controls, usage quotas and custom model training

Who it's for

AIRI lab serves architecture, interior, landscape and urban design studios that produce many concept images for competitions, pitches and client meetings. Its customer quotes include AUBE Conception, a French practice with teams in Paris and Shenzhen, and several Chinese design institutes. Enterprise plans for teams of three or more add bulk licensing, onboarding, commercial usage rights and cloud or on-premises deployment.

Team

The company was co-founded by architect Xavier Travert (CEO) and data scientist Feng Dayuan, known as Jacques Feng (CTO). It is headquartered in Shenzhen and runs global and European versions of the platform.