TL;DR

I led the frontend user experience for Active Theory’s internal AI hub, a custom tool designed to streamline studio-wide operations. Working closely with backend developers, I translated complex model integrations into a polished, intuitive web interface that mimics the familiarity of ChatGPT while surfacing deep agency intelligence. My focus was on shaping the entire user experience from high-level layout to micro-interactions to ensure that powerful AI functionality felt accessible and highly performant for every team member.

Impact

The platform revolutionized internal workflows by automating complex project intelligence and resource strategy. It transformed manual data searching into instant insights, allowing Producers to identify unbooked resources, estimate project costs based on historical data, and review specific client learnings for new scopes. For the Growth team, the tool acted as a strategic opportunity generator, providing data-driven recommendations for lead outreach and analyzing historical pitch success factors to maximize contract conversion rates. This shift replaced fragmented data silos with a unified point of truth that empowered the team to make faster, more informed decisions.

Description

Developed as an "Agile AI Assistant" ecosystem, the project was built to move the agency toward a chat-based intelligence hub. The platform serves as a "Digital Oracle" for the agency, acting as a centralized "collective memory" where years of project learnings, resource strategies, and sales intelligence are instantly accessible through natural conversation. Much like a bespoke version of Gemini or ChatGPT, the interface features specialized, custom-trained models tailored to different internal needs: Producers query timelines and milestones, Developers access an AI co-pilot for debugging and documentation, and Growth leads generate prioritized outreach lists.

The interface features a polished, interactive design that allows users to switch seamlessly between models depending on their specific task. Whether a team member is looking for the latest learnings from a specific client or needs help understanding how an internal code framework handles routing, the tool provides contextual, data-driven answers. By bridging the gap between sophisticated back-end task execution and front-end accessibility, the platform has become an essential "internal disciplines translator," ensuring that the entire agency can leverage its collective data to win and execute projects more efficiently.