KEND AI – Dual-Mode Interface for AI-Native Architecture
KEND AI – Dual-Mode Interface for AI-Native Architecture
Client:
Client:
Poliark-KEND AI
Poliark-KEND AI
Tools:
Tools:
Figma, Flowmapp
Figma, Flowmapp
KEND AI – Dual-Mode Interface for AI-Native Architecture
Designing for Architects in the Age of AI
KEND AI is a speculative UI/UX concept for an AI-native architectural tool. It introduces two experience modes:
🖋️ Conceptual Mode – for LLM-supported sketching and early idea generation
📐 Professional Mode – for precise architectural modeling with BIM-level parameters
As both an architect and a digital designer, I explored how product thinking and generative design workflows can converge into a simple, dual-mode interface.
The Challenge
Architectural software tends to be either too technical or too abstract. KEND aims to create a hybrid tool that is accessible for early-stage ideation, yet powerful enough for precision modeling.
My Approach
I designed a speculative interface that scales across personas — from architecture students to urban consultants — while keeping the UI minimal and prompt-friendly. I also integrated future-forward features like style transfer, zoning-based code checks, and LLM-based assistance.
The Challenge
Architectural software tends to be either too technical or too abstract. KEND aims to create a hybrid tool that is accessible for early-stage ideation, yet powerful enough for precision modeling.
My Approach
I designed a speculative interface that scales across personas — from architecture students to urban consultants — while keeping the UI minimal and prompt-friendly. I also integrated future-forward features like style transfer, zoning-based code checks, and LLM-based assistance.
The Challenge
Architectural software tends to be either too technical or too abstract. KEND aims to create a hybrid tool that is accessible for early-stage ideation, yet powerful enough for precision modeling.
My Approach
I designed a speculative interface that scales across personas — from architecture students to urban consultants — while keeping the UI minimal and prompt-friendly. I also integrated future-forward features like style transfer, zoning-based code checks, and LLM-based assistance.
Dual-Mode UX Flow
The UI splits into two complementary user journeys:
Conceptual Mode:
Prompt-based interaction → AI output → refine or transfer style → 3D export/transfer to pro modeProfessional Mode:
Location entry triggers code compliance → parameter input (FAR, sqm, etc.) → schematic model → refinement → BIM export
Each mode is visually unified through shared design tokens, layout logic, and interaction principles.
Product Roadmap
The case also includes a speculative product roadmap with future features such as:
Live Collaboration
Code Compliance by Location
Style Transfer from Image
Material Properties AI
Cost Estimation AI
And long-term visions like:
LLM Assistant for Urban Design
Project History Tracking
Sustainability Scoring
Dual-Mode UX Flow
The UI splits into two complementary user journeys:
Conceptual Mode:
Prompt-based interaction → AI output → refine or transfer style → 3D export/transfer to pro modeProfessional Mode:
Location entry triggers code compliance → parameter input (FAR, sqm, etc.) → schematic model → refinement → BIM export
Each mode is visually unified through shared design tokens, layout logic, and interaction principles.
Product Roadmap
The case also includes a speculative product roadmap with future features such as:
Live Collaboration
Code Compliance by Location
Style Transfer from Image
Material Properties AI
Cost Estimation AI
And long-term visions like:
LLM Assistant for Urban Design
Project History Tracking
Sustainability Scoring
Dual-Mode UX Flow
The UI splits into two complementary user journeys:
Conceptual Mode:
Prompt-based interaction → AI output → refine or transfer style → 3D export/transfer to pro modeProfessional Mode:
Location entry triggers code compliance → parameter input (FAR, sqm, etc.) → schematic model → refinement → BIM export
Each mode is visually unified through shared design tokens, layout logic, and interaction principles.
Product Roadmap
The case also includes a speculative product roadmap with future features such as:
Live Collaboration
Code Compliance by Location
Style Transfer from Image
Material Properties AI
Cost Estimation AI
And long-term visions like:
LLM Assistant for Urban Design
Project History Tracking
Sustainability Scoring
