
At Meta, posts helped teams share progress and priorities. But preparing weekly updates meant piecing together information from calendars, documents, and conversations—time taken away from the work itself.
Rather than introduce another communication tool, I chose to use AI to make participating in Meta's existing posting culture easier.
I designed a Claude- and MCP-powered workflow that synthesized existing work context into structured weekly updates.
Individuals could then refine the message, verify references, and decide what to share.
Calendars, tasks, documents, and collaborators provide work context.
AI synthesizes context into structured updates with relevant references.
People refine the message, verify details, and approve sharing.
A shared post keeps teammates informed, aligned, and connected.
Each update combined a weekly synopsis, workstream priorities, supporting documents, and relevant collaborators in a format colleagues could quickly scan.
What began as a design-team workflow spread to engineering, product management, and research—without a formal mandate.
In my experience, weekly updates went from roughly 45 minutes of preparation to 5 minutes with AI, including review and edits—an estimated 89% time savings.

Scaling beyond the original team meant making experience standards actionable across products and organizations.
A reusable experience template, cross-functional alignment, and ongoing design leadership helped translate those standards into new implementations—without compromising the experience's core.
Codified required elements and allowable variations across products, policies, and entities.
Directed conceptual design, cross-functional alignment, research validation, and execution.
Co-defined the strategic narrative that helped resolve launch barriers, including for Instagram.
Leading the experience from MVP through subsequent refinements taught me to balance customer needs, business priorities, and technical constraints without losing sight of usability.
That perspective shaped how I approach product challenges today: establish direction, make deliberate tradeoffs, and use evidence to continually improve the experience.
A useful product isn't defined by how much it can do, but by how confidently people can use it.
Technology can make communication easier without necessarily bringing people closer together.
This experience reinforced my belief that AI should do more than accelerate tasks. It should reduce the friction that prevents people from participating and connecting.