The specific visuals and details for this project are confidential.

Designing for Factual Integrity

UX for a critical internal platform for Google’s Knowledge Graph

Google | Interaction Designer | 2022 - 2024

What is it

The accuracy of Google's Knowledge Graph and Knowledge Panels is a matter of public trust.

While automated systems form the first line of defense, a sophisticated human-in-the-loop process is essential for verifying high-stakes information during sensitive events like elections, health crises, and breaking news.


My role was to own the end-to-end UX for "Project Veritas" (a pseudonym), the internal platform at the heart of this process.

The challenge was to take the existing platform, which lacked dedicated design ownership, and evolve it into a system that would empower a global team of human curators to work with speed, precision, and scale, directly safeguarding the information seen by billions.

What I did

As the sole UX designer, I was responsible for the tool’s design strategy and execution. My core task was to design for the transformation of the platform from a manual operation into a highly efficient, intelligent workbench for expert curators.

LLM/AI integration

Conceptualized and created designs incorporating LLM/AI.

End to end workflows

I designed workflows for task creation, assignment, etc

Design language

Created a comprehensive system of colors, typography, and interaction patterns where none existed, bringing consistency and predictability to the curator experience.

Universal taskbar

This core component, features AI-powered notifications

Data visualization

Dashboards to allow managers to proactively address issues impacting quality and delivery timelines.

What follows is a high level representative walk through of the project without any specific details to preserve confidentiality.

A story of impact: How my work ensured accuracy

To understand the impact of my work, let's imagine a hypothetical scenario involving the city of San Francisco.

After: New intelligent workflow

  1. A user sees an outdated population number in San Francisco's Knowledge Panel

  2. They submit a correction via “Feedback”.

The feedback is instantly analyzed and automatically routed to the correct curator. Using the modular taskbar I created, the curator gets AI-powered suggestions and can verify the new data and approve the correction in minutes, not days.

This feedback would land in a general queue, where a manager would have to manually find the right curator and assign the task—a process that could take days.

After the manager has manually assigned an expert, the information is updated in the knowledge graph.

Before: A manual, slow fix

  1. A user sees an outdated population number in San Francisco's Knowledge Panel

  2. They submit a correction via “Feedback”.

Outcome & Impact

Contributed to a significant increase in the number of tasks completed.

Modular taskbar component reduced developer integration from weeks to hours.

The design patterns I created were adopted as the new standard for development on the Search resultsThe UX foundations I established allowed the platform to scale globally and continue delivering value long-term.