GenAI Chat for Industrial Support
Enterprise AI tool to reduce support load and reuse internal knowledge
Global Fortune 500 industrial technology company · Enterprise-scale · 10,000+ internal users · Agency: Virtual Identity AG
Context
Role
Product & UX Designer
Timeline
2024–2025
Outcome
Scope
Constraints
NDA · Multi-region · Complex routing
Confidential Project: Some details have been anonymized to protect client confidentiality.
The Challenge
Users across multiple regions needed instant access to complex product information, but each region had different catalogs, languages, and compliance requirements.
Traditional search and static FAQs weren't cutting it. The business needed an AI-powered assistant that could route queries intelligently and seamlessly escalate to human agents when needed.
Process Overview
Understand
Mapped user journeys across regions. Identified friction in support escalation, product discovery, and cross-catalog search.
Decide
Prioritized progressive disclosure over all-at-once UI. Defined routing logic for AI vs. human handoff based on query complexity.
Design
Created conversation flows, response patterns, error states, and session management. Designed for transparency when AI can't help.
Deliver
Delivered complete design system with wireframes, interaction specs, and measurement strategy. Project continued after my departure.
Responsive interface across desktop, tablet, and mobile platforms
Design Decisions
- →Progressive disclosure over all-at-once: Instead of showing all capabilities upfront, the assistant reveals options contextually based on conversation flow, reducing overwhelm.
- →Transparent AI limitations: Clear messaging when the AI cannot help, with immediate human handoff options. Users trust honesty over frustrating loops.
- →Session continuity: Smart session management ensures users can pick up where they left off, even after timeouts or region changes.
- →Structured response hierarchy: AI responses use clear visual hierarchy with scannable headers, bullet points, and actionable next steps.
Design Artifacts — Illustrative Mockups
Visual mockups created to illustrate the interaction patterns and system behavior. Original production screens are protected under NDA.
Logged-In Welcome Screen
Not Logged-In Welcome Screen
AI Processing State
AI Answer - Structured Output
Error State - Variant 1
Error State - Variant 2
Human Agent Escalation
NDA-Compliant Visuals: All screens in this gallery are fictional UI recreations. The behaviors, logic, and flows reflect my actual design work, while the visuals have been intentionally anonymized.
Key Design Insights
Progressive Disclosure
Instead of overwhelming users with options, the assistant reveals capabilities contextually, suggesting next steps based on conversation flow.
Transparent AI Limitations
Clear messaging when the AI can't help, with immediate human handoff options. Users appreciate honesty over frustrating loops.
Session Continuity
Smart session management ensures users can pick up where they left off, even after timeouts or region changes.
Outcome
- —Complete design system delivered for conversational interface
- —Defined North Star Metric: Successfully resolved intents per active user
- —Created measurement framework for user satisfaction and AI effectiveness
- —Project continued into development phase after my departure
The projected impact targets 40% reduction in support tickets and 85%+ user satisfaction — metrics defined during the measurement strategy phase I led.
What This Project Demonstrates
- •Conversational UX and AI interface design
- •Enterprise-scale product thinking
- •Complex routing and state management design
- •Measurement strategy and success metrics definition
- •Working within NDA and compliance constraints
- •Cross-regional design considerations