Marketing

Automating weekly meeting recaps and action items

Use an AI assistant connected to meeting transcripts (e.g., Granola), Google Docs, and Slack to review the full call transcript, cross-reference running notes/work plans, scan relevant Slack channels/DMs, and draft a formatted weekly recap message with action steps, owners, and due dates for the team.

Why the human is still essential here

A human still reviews the transcript context, validates priorities/owners/dates, and edits the draft before sharing—ensuring accuracy, tone, and accountability stay human-led.

How people use this

Fireflies.ai recap to Slack draft

AI transcribes the weekly marketing meeting, extracts decisions and action items, and drafts a formatted recap message to your Slack channel for human review.

Fireflies.ai / Slack

Otter action items to Asana tasks

AI converts transcript action items into assigned Asana tasks with due dates, then compiles a weekly recap for the team to confirm before sending.

Otter.ai / Asana

Zapier-based weekly digest from Docs + Slack

An automation pulls context from Google Docs notes and recent Slack threads, then uses an LLM to draft a Monday recap/work-plan update as a Slack message.

Zapier / ChatGPT / Google Docs / Slack

Action items synced to Asana with owners and due dates

Fathom summarizes the meeting and Zapier creates or updates Asana tasks for each action item with an owner and due date pulled from the recap.

Fathom / Zapier / Asana

Client call summary pushed to HubSpot timeline

Avoma generates a client-call recap with follow-ups and automatically logs the note and related tasks to the appropriate HubSpot records for visibility across the team.

Avoma / HubSpot

Community stories (1)

LinkedIn

2 actual ways I used AI in my workflow this week.

2 actual ways I used AI in my workflow this week. One saved me 30–60 min on something I dreaded. One saved me 10+ hours and helped me do my job better.

Working out loud because I find actual examples way more useful than general "you should be using AI" advice. I'm not an expert, just figuring it out like everyone else.


I've been sorting experiments into 2 buckets:

1. Efficiency: what saves me time or kills manual work

2. Effectiveness: what helps me do better, more strategic work


One from each this week.


1. Automated weekly meeting recap (Efficiency)


→ Tools: Claude Cowork with connectors to Granola, Chrome extension for Google Docs, and Slack

→ Logic: Automate weekly comms with the team following team call with recap and work plan action steps (with due dates + owners)

→ Setup: Reviews the full meeting transcript (not just the AI summary), cross-references our running meeting notes and separate running work plan in Google Docs, scans certain Slack channels and DMs from the previous week, identifies the recap + action items with due dates per stakeholder, then drafts a formatted Slack message to our team channel. This could automatically post, but it's not 100% perfected yet and I still edit certain things, so it lands as a draft.

→ Scheduling: Every Monday, 5 minutes after our weekly call

→ Time saved: 30–60 min/week — 2–4 hours/month on a task I would have 100% passed to an intern if I had one


2. Interactive messaging wireframe (Effectiveness)


→ Tools: Claude Chat

→ Logic: MVP mock up messaging and hierarchy for a homepage rewrite to communicate visually with the team and avoid confusion

→ Setup: Chat with Claude to mock this up: creates an interactive HTML file you can see in the chat, download, and open in your browser (can also send to the team to do the same). Supplied it with the current website, other files about the company/messaging, what I wanted to change, etc.

→ Time reduced: Goodness, probably 10+ hours. Not mocking it up in Figma/Canva, the team is able to review faster, etc. The copy alone saved me multiple hours. I still edit and direct it very closely, but it gets to a solid MVP point to communicate direction and wordsmith from. I was able to rewrite the whole homepage and have a first draft to review with the team in an afternoon.

→ Notes: This one is SO cool. I was also able to create toggle options to show different messaging approaches in the same file. The file is editable too, so as you're discussing as a team you can edit the copy directly. Presenting work visually like this is something I feel strongly about.


I've been blown away with Claude lately. Fully on the hype train and the ChatGPT transfer crowd! What I really like is how much it doesn't just fully agree with me (have this in my settings too), and how fast it moves.


Anything I could be doing better on these two? Any cool ways you've been using AI I should try next? Learning out loud with everyone here!

AL
Ashley LewinMarketing Leader
Mar 5, 2026