AI Control Panel

---

Role: Lead Product Designer
Duration: 2 years (2023 to 2025)
Platform: Web App
My Responsibility: End-to-end product design from discovery to handoff

Tools: Figma · Jira

The goal of this project was to design a control panel that gives users transparency and control over AI-driven automation in healthcare operations.

The system allows users to adjust AI confidence thresholds for different models such as:

• Claim Denial Predictions – identifying which claims are likely to be denied,
• Patient Record Mapping – merging duplicate patient records, and
• Doctor Record Mapping – consolidating provider information.
   

When the AI’s confidence score exceeds the threshold, it automatically completes the task. If not, the task is routed to a human reviewer. This balance ensures both efficiency and accuracy.

To help users see the tangible benefits of automation, I designed a dashboard that visualizes:

• The total work completed by AI over a custom date range,
• Cost and time savings, and
• The equivalent full-time employee (FTE) savings.
   

The result is a clear, data-backed way for organizations to measure how AI contributes to their operational efficiency and return on investment.

Methods:

• Stakeholder interviews with AI ops leads and analysts
• Studied model performance reports to identify user pain points
• Analyzed logs to understand how threshold changes impacted workload
   

Key Insights:

• Non-technical users were hesitant to modify model parameters.
• Users cared more about outcomes (savings, accuracy) than technical metrics.
• Visual correlation between thresholds and results would improve trust.
   

🧠 “I just want to know — if I change this slider, how many tasks will AI take over?”

1. Control: Let users safely adjust automation thresholds.
2. Clarity: Show real-time impact of changes (AI vs. human workload).
3. Confidence: Make the relationship between threshold and outcomes transparent.

The product was divided into three main sections:

1. Dashboard Overview — High-level AI performance metrics:
 

• % of work auto-resolved by AI
• Total AI-driven savings (in FTEs & hours)
• Accuracy trend over time
• Custom date range filters
   

2. Model Management — Configure AI thresholds per model:
 

• Claim Denial Prediction
• Patient Record Mapping
• Doctor Record Mapping
• Future models could be added dynamically
   

3. Impact Analysis Panel
 

• Visualize cost savings and volume handled by AI
• Compare human vs. AI productivity
• Export insights or reports

• Partnered with data scientists to translate AI metrics into understandable KPIs.
• Collaborated with developers to integrate live threshold APIs.
• Defined QA checklist for accuracy, loading time, and responsiveness.
• Created a mini design system for dashboard patterns, chart modules, and sliders.

• Design: Figma, Axure RP
• Management: Jira, Confluence
• Methodology: Agile with iterative design and development cycles

• Conceptualization: Created wireframes and prototypes based on user feedback.
• User Testing: Conducted usability testing with coders and stakeholders to refine the design.
• Implementation: Worked closely with developers to translate designs into a functional product.
• Iteration: Incorporated feedback to improve features and workflows post-launch.

• Balancing diverse user needs: Resolved through iterative testing and stakeholder feedback.
• Ensuring compliance with medical standards: Achieved by embedding industry guidelines into the validation logic.

The control panel helped teams reduce manual workload by over 40%, improving operational efficiency and demonstrating clear ROI from AI automation.

Can provide on request