Overview
Designed and deployed a best-in-class web application enabling law enforcement to rapidly locate and recover stolen vehicles through SiriusXM's connected car technology, reducing recovery times and preventing secondary crimes.
Deliverables
Product Discovery
Research Automation
AI Powered Product Knowledgebase
Design-Engineering Bridge
Using agentic workflows to collaborate more and build a best-in-class stolen vehicle recovery solution.
Executive Summary
Our UX team leveraged AI-powered workflows to dramatically reduce time to market while enhancing collaboration across product management, engineering, and program management. By implementing intelligent automation through n8n workflows and custom GPTs, we transformed our traditional design research and development process into a streamlined, data-driven system.
If you only read one thing…
We built agentic workflows(n8n) that capture, process, and share all our product insights
We improved collaboration and clarity by aligning UX, product, and engineering
Achieved 60% faster research synthesis and 100% faster UX artifact creation
Implemented real-time design system + Material UI 3 sync, cutting handoff errors by 50%.
Success from treating AI as augmentation, not automation—maintaining human creativity.
The Problem: Manual analysis of customer discovery data, repetitive creation of design artifacts, and inconsistent communication between product, design, and engineering teams was prolonging our time to market.
Our Approach:
We developed two complementary solutions.
Improve our time-to-insight and product clarity.
AI Research Workflow: Automatically captures customer interviews, call logs, and competitive analysis to generate a custom GPT that creates personas, service design artifacts, identifies technical risks, and drafts prototype requirements.Improve our Production Process and Engineering hand-offs
Design-Engineering Sync: Uses vibe-coding to maintain reliable connection between our design system and Material UI 3 component libraries, eliminating handoff bottlenecks.
Implementation
Phase 1: Research Automation
Set up data connectors for interview transcripts, contact center logs, and competitive analysis. Configured n8n workflow to continuously process new inputs and train custom GPT on project-specific context.
Phase 2: AI Assistant Integration
Trained custom GPT to understand our product domain, user personas, and technical architecture. Established validation protocols comparing AI-generated artifacts against manual versions.
Phase 3: Design-Engineering Bridge
Configured vibe-coding workflow to monitor design system and component library changes, automatically flagging discrepancies and maintaining Material UI 3 alignment.
Results
Quantitative Impact:
Research synthesis time: 60% reduction
Service design artifact creation: 100% faster
Technical risk identification: 35% improvement
Design-to-engineering handoff errors: 50% reduction
Qualitative Benefits:
The shared AI assistant created common language between design, product, engineering, and program management teams. Cross-functional stakeholders could query the same knowledge base, leading to more aligned decision-making. Team members gained bandwidth to focus on strategic design thinking and innovation rather than routine documentation tasks.
Key Learnings
Success Factors:
Data quality was crucial—AI effectiveness was directly proportional to input data comprehensiveness
Cross-functional buy-in required early involvement and continuous iteration based on team needs
Human-AI collaboration worked best when AI augmented rather than replaced creative expertise
Challenges Overcome:
Initial skepticism addressed through gradual introduction and transparent capability demonstration
Data privacy concerns resolved through robust governance protocols and security team collaboration
Technical complexity managed through platform engineering partnerships and team training
Conclusion
AI-enhanced workflows enabled us to accelerate time to market while maintaining design quality and strengthening cross-functional partnerships. Success came from positioning AI as intelligent augmentation that amplifies human creativity rather than replacing it.
The collaborative approach we developed creates a sustainable model for rapid, high-quality product development. For organizations considering similar initiatives, focus on clear objectives, data quality investment, team training, and human-AI collaboration rather than replacement.
Future opportunities include usability testing integration, A/B testing analysis automation, real-time accessibility compliance, and predictive analytics for user behavior forecasting.
