Crits scattered across meetings, docs, and chats? Here's a workflow that actually connects them.
Connecting Google Workspace and Figma | Config 2026 · Jenny
21 min total·Actually worth watching closely: ~6 min·3 must-watch clips
- 0:12 – 2:10Listen
Start by admitting fragmentation can't be eliminated
The opening lays the reality of the design process out flat: conversations in meetings, decisions in documents, feedback in chats, and nobody should expect it all to consolidate into one tool. The whole argument rests on connecting the tools that matter rather than eliminating fragmentation.
The goal isn't to herd everyone into a single tool, it's to make ideas and discussions reachable across tools so work keeps its continuity.
The speaker standing and making an argument, with essentially nothing on screen you need to watch. Fine to listen while doing something else.▶ Jump to 0:12Speaker · Jenny - 2:10 – 4:09Listen
Make the design process itself visible
Moves from collaboration pain to one concrete practice: carve out a designer-owned section inside the product requirements doc and let design principles, executive steerings, cross-functional feedback, and research insights accumulate there.
Visibility into the design process and the design thinking matters as much as the files you ship. Give it a fixed home where it can build up over time.
An organizational practice that comes across fully in words; the screen adds no information.▶ Jump to 2:10Speaker · Addison Rodomista - 4:09 – 5:52Watch
Design links become cards in the doc
A live demo of pasting a design file link into Google Docs, where it turns into a native-looking card that expands on hover with file metadata and a thumbnail, replacing a long string of a URL nobody can read.
A design reference in a doc goes from being a link to being previewable context, so readers don't have to jump out to confirm which file it is.
The value here is entirely on screen: the moment the link becomes a card and the thumbnail pops up on hover is exactly what a text summary loses.▶ Jump to 4:09Speaker · Addison Rodomista - 5:52 – 7:16Watch
Bring the canvas into the video call
The demo then embeds the full interactive canvas directly into Google Meet: the presenter pans and zooms freely and shows microinteractions while staying aware of how the room is reacting, and the others can enter the canvas and work in it too.
The crit stops being a one-way slideshow and becomes everyone working hands-on in the same live canvas.
The feel of driving the canvas and of several people joining in at once only comes across visually; on audio alone it sounds like ordinary screen sharing.▶ Jump to 5:52Speaker · Addison Rodomista - 7:16 – 10:20Listen
The meeting ends, the context stays
What happens after the crit: the discussion from the call becomes persistent, searchable chat, later design changes and tags get pushed to the right people automatically, and you can reply to comments from your phone to keep things moving.
The real value of a crit lands afterwards. Leaving the discussion as a searchable record matters more than saying more in the room.
This stretch strings a workflow together verbally; the visuals contribute little.▶ Jump to 7:16Speaker · Addison Rodomista - 10:20 – 14:18Skim
How notifications avoid becoming noise
Down into the detail of notification design: the system knows whether you're actually looking at that file right now and decides whether to interrupt you accordingly. Then it pulls up to the platform view of how this integration holds up across large teams.
The hard part of connecting tools isn't whether you can push a message, it's restraint. Notifications have to sense what the user is doing or they become a new source of distraction.
Mostly platform-vision slides. Catch the short argument about notification awareness and you can fast-forward through the rest.▶ Jump to 10:20Speaker · Addison Rodomista - 14:18 – 17:40Skim
From chatting to actually doing the work
Turns to AI and traces its arc in the workplace: from chatbots that speed up a conversation, to autonomous multi-step agents that complete real tasks, with the point that an agent's true bottleneck is not having context across apps.
When an agent can't run a workflow well, it usually isn't the model. It's that the company's data is scattered across apps where the agent can't see it.
Narrated over a diagram of the evolution stages; the diagram reads in a glance, and the substance is in the few sentences of conclusion.▶ Jump to 14:18Speaker · Chloe Fagan-Tucker - 17:40 – 21:01Watch
A context interface for developers
The closing announcement is the Workspace MCP server for developers: natural language prompts in place of the existing API call structure, with skills across mail, Drive, Calendar, Chat, and people directories plus universal search, currently in developer preview and being built out with early access partners.
To wire enterprise context into an app you build, the bar drops from reading an API reference to writing one natural language prompt, and you can start today.
The capability list and the entry point are both on screen; developers will want to look at exactly which skills are available and where to get in.▶ Jump to 17:40Speaker · Chloe Fagan-Tucker