Loading thumbnail…
Agentic Skills: Build Once, Use Over and Over
What agent skills are, how to write ones that work, where to find skills others have built, and how to chain them into pipelines that check their own work.
Watch on YouTube
A living home for our monthly AI webinars and tool updates — a fresh session every two weeks, recorded for anyone who missed it, plus the tools your team has access to and real show-and-tell from people using them.
A retrospective archive of every AI webinar we've held. New session every month — older ones stay here for anyone who missed it or wants a refresher.
What agent skills are, how to write ones that work, where to find skills others have built, and how to chain them into pipelines that check their own work.
How loops in Workspace Studio flows run the same steps across a whole list — build an action once and Studio works through every row, file or result unattended.
A hands-on walkthrough of chaining NotebookLM, Gems and Google Workspace Studio together into one workflow.
How leading brands are rethinking content creation with AI — from process and team setup to the output itself.
Using AI to run market research — faster insight gathering, synthesis, and analysis.
Creating content at scale with AI — the emerging "content engineer" role and the workflows behind producing more, faster.
Where agentic AI is heading for marketing teams — what autonomous, tool-using AI means for how we work, and where it's already taking hold.
An introduction to building automated flows in Google Workspace Studio — the basics, set up step by step.
The REAP framework for validating research — a structured way to pressure-test findings before you rely on them.
Each month we walk through the AI tools your team has access to — what's new, what changed, and how to get the most from them. We often invite people to show-and-tell how they actually use them.
This month's briefing ran through a fast-moving market and the tool changes that matter for how we work. On Google's side: Gemini now builds real, editable decks in Slides from your audience and goal, "Fill with Gemini" landed in 11 more languages, Vids turns a deck into a narrated AI-avatar video, and Gemma 4 runs fully offline on a Pixel 10. On models: GPT-5.6 arrived with built-in design judgment and strong token efficiency, Fable 5 returned as the top all-rounder, and the running theme was picking the right model for the task as agent token costs climb. Plus a wave of agent features — ChatGPT Work mode, Claude's Tag for Slack and Reflect tab, Meta's Muse — and our own video pipeline cutting cost per video from about £6 to 23p.
Gemini now generates editable, on-template slide decks in Google Slides from your audience and goal — then Vids turns them into narrated AI-avatar videos.
ChatGPT Work mode and Claude's new Tag for Slack and Reflect tab embed AI where you already work — and start suggesting what to automate.
Open Montage, the AI video system we're testing, hands storyboarding, scripting and design to Claude and GPT — cutting cost per video from about £6 to 23p.
Ask Gemini in Google Chat, loops in Workspace Flows, NotebookLM's new exports, Perplexity's Comet browser, ChatGPT agents and Claude's live artifacts.
What's scheduled next — webinars and tool updates each run monthly, alternating so there's a fresh session every two weeks.
A hands-on, build-along session. Bring one repetitive task from your week and we'll turn it into a working skill together — start to finish, live.
The next monthly walkthrough of the tools your team has access to — what's new, what changed, and another round of show-and-tell.