Signals Dashboard

Designing a flexible, AI-powered tool to transform high-frequency energy data into actionable insights for operations managers.

Sales needed a demo tool to showcase Verdigris’ energy intelligence. But through customer interviews, we uncovered a deeper pain point:
• Operators receive alerts only after equipment fails.
• They wanted early warnings, AI analysis, and actionable recommendations.

• Conducted interviews with facility operators and directors.
• Key finding: alert fatigue was a major blocker. Customers needed fewer but more meaningful signals.
• Shifted focus audience from sales → operations managers.

1. User Flow & Information Hierarchy
  • Placed granular, high-frequency data at the core.
  • Drafted flows for anomaly → symptom → signal → recommendation.

2. Cross-functional Workshops
  • Collaborated with PMs, data scientists, engineers, and GTM team.
  • Defined wireframes together and aligned on MVP scope.

3. AI-Accelerated Design
  • Trained AI with our design system.
  • Turned hand sketches into polished mockups instantly.
  • Enabled PM/engineers to fork design ideas → validation loop compressed from 2 weeks to hours.

4. Testing & Iteration
  • Usability sessions with operators validated value proposition.
  • Explored features like integrations, deep-dive research, and interactive configuration.

A proactive anomaly detection dashboard that:
  • Detects symptoms (e.g., current imbalance, short cycling).
  • Combines them into signals requiring action.
  • Provides AI-driven recommendations with option for deeper research.
  • Balances trust, clarity, and actionability to reduce noise.

• Internal stakeholders (sales, PM, engineers, marketing) reported faster alignment through AI-accelerated design.
• Pilot customers and domain experts in Design Partnership Panel validated the approach.
• Anticipated as a “killer product” in the Verdigris line, launching September.

• Lead Designer driving product vision, user experience, and design process.
• Mentored a junior designer & guided a design researcher.
• Facilitated cross-functional collaboration with ML, data science, engineering, and GTM teams.
• Established design principles around actionable signals vs. raw alerts.