Google Economic Impact

Insight

As part of this end-to-end project, I focused on a series of improvements to the economic impact report website. The first of which was integrating an AI-powered search function, aimed at refining how users access and retrieve information from the report.

Followed by creating an AI-powered agent for more personalized assistance with report materials and a site audit to bring visual improvements to the total platform experience.

What I did

As the experience designer and AI point person on the creative team, I championed the phases of developing the AI search and AI chat agents. Researching AI capabilities and affordances, building iterations of interfacing alongside the visual designer, working closely with the development team to identify technical feasibilities under the given timeline, and presenting recommendations to stakeholders.

I also made high contributions in the site audit through my evaluation and organization of recommendations, collaborating with the creative director and visual designer to refine the illustration of visual changes to present to stakeholders.

Building AI search

Leveraged user archetypes, comparative analysis, and AI documentation to design concepts aligned with both user needs, Google's visual standards, and their technological infrastructure.

Collaborated closely with the development team to explore Vertex AI's affordances, prototyping and refining the search functionality to optimize data retrieval. This included examining various data types for the AI model and providing stakeholders with recommendations on the optimal dataset to use for generating the best search results.

Result

An MVP feature that leveled up the platform's information retrieval capabilities and set the stage for future AI-driven enhancements. Production expected in 2025.

Mapping an AI agent

While similar to the approach of AI search, this feature delved deeper into enabling users to generate more detailed insights and explanations of report materials.

Continuing the cross-functional collaboration with the development, delivery, and creative teams, we selected an alternative dataset to better align with the features generative goals, developed conversational mapping to establish guardrails against hallucination, and proposed visual treatments for the agent's interface.

Result

An MVP feature that advanced the platform's ability to offer deeper user engagement and more comprehensive support, paving the way for more personalized interactions with site material. Production expected in 2025.

Strengthening the report

Leveraged data analytics, heuristic evaluation, and user archetypes for measuring performance.

Identified areas to improve through the above methods and build a robust set of platform recommendations. Followed up with prototype mockups to illustrate recommendations visually.

Impact

The site was launched with treatments that made prominent features shine brighter like the interactive map and gave additional access points to informative pages like U.S. Impact. Ultimately delivering a highly informative platform revered by Sundar Pichai himself (CEO of Google).

See the website live