AI systems & spatial interfaces
for enterprise decisions
Impact
- Simplified AI architecture 10+ → 3 agent types
- Designed SuperAgent to orchestrate the AI ecosystem
- Built spatial interfaces for Apple Vision Pro
Project
overview
Background
Aily Labs was building an AI-driven decision intelligence platform where multiple AI agents analyzed data, generated insights, and proposed actions.
Challenge
The system had grown organically and began to show structural challenges. The platform contained over 10 different agent types with overlapping responsibilities, creating complex orchestration flows that were difficult to understand, scale and maintain.
Objectives
- Simplify the underlying AI architecture
- Design product experiences that translate complex AI outputs into clear decisions
- Generate trust
- Innovate in how we interact with data
Starting Point
- Ad hoc agent system
- Over 10+ agents with no clear structure
- System difficult to scale
- No hollistic connection between agents
Rethinking the AI Agent Architecture
New structure
The first step was analyzing how agents interacted across the system. The architecture was redesigned into three core agent types, each with a clear responsibility in the decision pipeline.
Super Agent
The Super Agent acts as the orchestration layer between agents and the user.
- Unified insights
- Contextual recommendations
- 360° view of the problem space
- Proactive decision-making tool
One of the main design challenges was translating AI outputs into interfaces that people could trust and understand.
Designing for Vision Pro
The future of data analisys
As part of the exploration of future interfaces, we designed spatial data experiences for Apple Vision Pro.
How spatial computing expands data interaction
001
Data beyond the screen
Removing the constraints of traditional displays allows us to visualize multiple datasets simultaneously, making comparison and pattern recognition significantly easier.
002
Immersive analysis
Complex AI insights could be explored in spatial layouts, helping users understand relationships and trends that are difficult to perceive in flat interfaces.
003
Collaborative decision spaces
Teams could enter the same virtual environment and interact with data together, turning analysis into a shared, real-time decision-making experience.
Sitemap
Social
© 2026 Carlota Antón. All right reserved.