<aside> <img src="/icons/arrow-left-basic_lightgray.svg" alt="/icons/arrow-left-basic_lightgray.svg" width="40px" />

Back to Main Portfolio

</aside>

This case study is under construction (011326)

<aside> <img src="/icons/arrow-northeast_orange.svg" alt="/icons/arrow-northeast_orange.svg" width="40px" />

Click here to check the final interactive prototype.

</aside>

1. Overview

High-stakes decision intelligence. Result: 9.5/10 Executive Confidence Score.

The Challenge

In an autonomous agent workflow, "human-in-the-loop" oversight was becoming the bottleneck. Executives (Sam) were "rubber-stamping" $50k+ decisions because risks were buried in dense tables, while Ops Leads (Maya) were slowed down by excessive "mouse mileage," capping daily throughput.

• The Outcome: We pivoted from a generic dashboard to persona-specific "Decision Modes," increasing Executive Confidence Score from 6/10 to 10/10 and achieving 100% Slack Deflection for request tracking.

The Outcome

We pivoted from a generic dashboard to persona-specific "Decision Modes," increasing Executive Confidence Score from 6/10 to 10/10 and achieving 100% Slack Deflection for request tracking.

The Setup

• Role: Lead Product Designer (Strategy + UX/UI + Frontend)
• Team: 1 PM, 2 Eng, AI Agents (Figma, Google Antigravity for Agentic Dev, Google AI Studio for fast prototypes)
• Timeline: 3 Weeks

2. Strategy & Solution (The Process)