Interfaces are changing radically. After decades of clicking buttons and menus, we’re entering the era of conversational interfaces: chatbots, copilots, AI agents that understand natural language.
Traditional GUI rules don’t work for AUIs (Agent User Interfaces). New challenges: how to build trust in a non-deterministic system? How to manage expectations when AI can make mistakes?
In Italy, no agency has specialized in this field. We have, developing AI Agents for our clients and learning what works on real projects.
Sales Copilot
B2B company with AI assistant for sales reps. The first prototype, developed without UX, had a problem: nobody used it (12% adoption).
Problems identified: Generic chat box without context. Sales reps didn’t understand what to ask, didn’t trust the responses.
Solution: Proactive suggestions, one-click actions, source transparency, ability to correct the AI.
A New Paradigm
From GUIs to AUIs (Agent User Interface)
Our Frameworks
Proprietary methodologies born from experience on real projects
Conversational UX
Trust Design
Hybrid Interface
Design Patterns
AI Agent Onboarding
- • Personality introduction
- • Initial expectations
- • Guided interactions
- • Complexity growth
Conversational Commerce
- • Product discovery chat
- • Conv. recommendations
- • Assisted checkout
- • Proactive post-sale
Support Agent
- • Automatic triage
- • Knowledge retrieval
- • Seamless escalation
- • Feedback loop
Copilot Interfaces
- • Contextual suggestions
- • Approvable actions
- • Collaborative edits
- • Learning from corrections