Menu
Book a free call CV
Designing Hexagon’s Unified AI Assistant Guidelines
Hexagon, a global leader in manufacturing, engineering, and simulation software, sought to unify its AI assistant experience across varied business units.

The challenge
Create unified AI assistant guidelines that deliver a consistent UX, maintain brand cohesion, and meet regulatory compliance requirements (EU AI Act), while remaining flexible enough to adapt across products.
Read Guideline Report
Company
Hexagon

Product
Internal AI Assistant Guidelines for Designers and Developers
Deliverable
PDF, SharePoint, and internal web page.

Timeline
4 weeks (Discovery to Delivery)
My Role
AI expert consultant.
Team
Collaborated closely with a Design System Architect, UX Researcher, Content Strategist, and AI Product Manager to ensure strategic alignment and cross-team integration.
MY RESPONSABILITIES

Served as AI Expert Consultant, bridging advanced AI and design practice by translating complex machine learning and generative AI capabilities into actionable, regulation-compliant UX/UI guidelines.

MY ROLE

Ensured guidelines were technically feasible and regulation-compliant.

Provided strategic advisory on best-practice AI implementation approaches for diverse product needs.​

Bridged the gap between engineering and design teams to streamline implementation.

Maintained both technical rigor and design excellence throughout guideline development.

Translated AI technical capabilities into actionable design requirements for UX/UI teams.

Impacts
  • 84% product teams adopted guidelines within 3 months
    Consistency achieved! Established unified approach across Hexagon's software ecosystem through Nova design system integration
  • Scalable Framework
    Successfully developed a scalable framework, enabling scalable AI across products with different maturity levels
  • Trust Mechanisms
    Compliant with EU AI Act & ISO/IEC 42001:2023 transparency standards.
    Implemented clear traceability patterns, helping users distinguish AI-generated content
  • +30% perceived product quality & brand alignment
    Reinforced Hexagon's positioning in precision and trustworthy AI. Product teams actively using the living documentation for implementation guidance
//
//
//
//
//
CHALLENGES

Design System Integration

Fragmentation Risk

New AI patterns to align with the existing HxGN design system infrastructure.

Trust Requirements

In high-precision industries, users need clear differentiation between human-generated and AI-generated content.

User Diversity

The guidelines needed to serve everyone from field technicians to data scientists, each with distinct needs.

Different Levels of Maturity

Different products had varying levels of AI sophistication, from basic tooltips to autonomous agents.
Without standardized guidelines, AI assistant implementations across different product teams could become inconsistent, leading to user confusion.

ORG GOALS

Transform Hexagon's diverse software portfolio into a cohesive ecosystem with intelligent, trustworthy AI assistance that strengthens our position as the leader in precision digital reality solutions.

Access intelligent, context-aware assistance that enhances productivity and decision-making quality in high-precision professional workflows—from field technicians to data scientists—with transparent AI involvement, maintained human control over critical operations.

USER GOALS

DESIGN GOALS

Create a flexible, regulation-compliant AI design framework that accommodates products at varying maturity levels, integrates seamlessly with existing Nova design system infrastructure, and establishes trust through transparent AI attribution and human-in-the-loop patterns.

PHASE 01 Foundations

I guided the team to establish core concepts and defined the AI assistant's narrative character:
  • Mapped AI technologies (ML, Deep Learning, Generative AI, Agentic AI) to Hexagon workflows and value
  • Defined a three-stage maturity framework: Guide, Interactive Partner, Autonomous Partner, to scale design for product readiness
  • Developed a unified assistant personality embodying expertise, credibility, and approachability
  • Aligned foundational concepts with EU AI Act transparency, accountability, and governance principles
See Maturity Framework
Maturity Stage Framework
Rather than forcing all products to adopt the same AI features, the guidelines defined three clear stages that products could progress through based on their technical capabilities and user needs:
  • Stage 1 (Guide)
    Contextual help, tooltips, and smart lookups powered by ML and contextual lookup
  • Stage 2 (Interactive Partner)
    Conversational UI with generative AI capabilities (LLMs, Deep Learning)
  • Stage 3 (Autonomous Partner)
    Proactive, multimodal interactions with agentic AI
3 Stage maturity Framework: Guide, Interactive Partner, Autonomous Partner—as detailed in Hexagon’s official AI Assistant UX/UI Guidelines (Chapter 2). This framework guided tailored UI patterns, personality consistency, and traceability design aligned with core AI capabilities and brand values.

PHASE 02 Initial Delivery

Delivered foundational design documentation fully integrated with Hexagon's Nova design system:
  • Created foundational design patterns integrated with Hexagon’s Nova design system
  • Delivered UI components tailored by maturity stage: tooltips, embedded chats, voice interfaces
  • Established traceability frameworks for clear AI content attribution with visual indicators and user controls
  • Published living documentation on Zeroheight for real-time versioning and team access

PHASE 03 Comprehensive Guidelines

Released focused UX/UI guidelines for priority use cases:
  • Expanded guidelines to cover multimodal interactions (voice, text) and conversational flows
  • Collaborated with Marketing, Legal, and Product teams to validate accessibility, compliance, and alignment
  • Conducted iterative reviews incorporating multi-stakeholder feedback

PERSONALITY

Created a consistent AI assistant personality that remained recognizable across all implementations while allowing flexibility in behavior:

  • Expert
  • Credible
  • Approachable

  • Clear
  • Concise
  • Confident

TONE

INTERACTION STYLE

  • Context-aware
  • Guiding
  • Purposeful

PHASE 04 Validation & Rollout

  • Supported adoption via training sessions and ongoing documentation updates
  • Monitored uptake—achieved 84% guideline adoption within 3 months across product teams
  • Gathered feedback for the continuous evolution of AI interaction patterns

Training & Enablement

Measurable acceptance criteria for the rollout readiness of the AI Assistant Guidelines, ensuring that the project and teams were prepared for successful adoption and deployment:

  • Minimum of 3 training sessions/webinars conducted with product teams and designers.
  • Over 80% of targeted users complete training with positive feedback (average satisfaction score ≥ 4/5).

  • Full integration of AI Assistant UI components with Hexagon’s Nova design system verified by Design System Architect and engineering leads.
  • Zeroheight documentation is linked with development pipelines and accessible by all implementers.

Integration & Tooling

Pilot Implementation

  • At least two products successfully implement AI assistant features using the guidelines in a pilot phase.
  • Pilot feedback collected and yielded ≥ 75% positive user evaluation on trust, usability, and brand consistency.

//
//
//
//
KEY TAKEWAYS

Traceability Builds Trust
Clear visual indicators for AI-generated content are essential for professional, high-stakes environments.

Maturity Models Enable Flexibility
The three-stage framework lets products adopt AI features appropriate to their readiness level.

Living Documentation Accelerates Adoption
Using zeroheight for real-time, versioned guidelines enables teams to access the latest patterns immediately.

Design System Integration is Critical
Aligning AI guidelines with the existing Nova design system ensures consistency and reduces implementation friction.

Made on
Tilda