Google Workspace AI Dashboard

People live inside the Google ecosystem all day, but the tools don't help them manage attention, priorities or mental load. The brief was open: how might AI genuinely improve the digital work experience across Google Workspace, without just adding more noise?

We ran a full Design Thinking process: desk research on AI at work and remote productivity, netnography on real user reviews and forums, and a survey of 50 intensive Google-ecosystem users. The numbers were clear:

The recurring pattern was notification fatigue and cognitive overload, and a reactive culture instead of strategic planning, the ecosystem acts as an 'external brain'. Tool by tool: Meet is too minimal for complex contexts; Calendar doesn't fail technically, it fails because it assumes an organised habit many users don't have; Gmail is powerful but pushes people into reacting instead of working. The core insight: the problem isn't technical, it's cognitive and behavioural.

So instead of adding features, we designed a layer that reduces load. We synthesised needs into personas and an empathy map, then prioritised ideas with MoSCoW and a utility/viability matrix. The concept: a visual, customisable dashboard that connects all of Workspace, with a compact AI assistant, smart event and task prioritisation, meeting transcription and summaries, error prevention (duplicate files, unsent attachments) and a cognitive-load indicator, always with human approval.

I focused on the information architecture, data visualisation and the compact AI assistant, presenting a lot of information without adding to the overwhelm the research had surfaced.

This project sharpened my research muscle: turning desk research, netnography and 50 survey responses into a clear, defensible problem, and resisting the urge to solve it with more features. The strongest insight came from reframing a 'technical' problem as a human one.