007 AI for Operators

Chief of Staff at Order.co, review of Granola.ai, 9 links

Hi there, 

Welcome back to AI for Operators for our James Bond edition (just kidding, but we do have a review of a sneaky AI tool)!

This week, here’s what we’ve got:

  • The Operator: Ali Sardinia, Chief of Staff at Order.co, walks us through her workflow to create beautiful, effective presentations in a fraction of the time.

  • The Review: Granola.ai - the controversial meeting notetaker

  • The Links: 9 links, including controversial acquisitions, AI web browsers, and the inside scoop from a departed OpenAI employee

The Operator

Ali Sardinia, Chief of Staff at Order.co

Ali is the Chief of Staff at Order.co. Previously, she worked as the Chief of Staff at Managed by Q.

In this episode, Ali showed how she uses Plus AI to automate presentation creation, saving her hours every week and improving the quality and design of the output.

  • Deck hours slashed: Plus AI auto-remixes drab slides into brand-aligned layouts inside Google Slides, trimming 3-4 hours from the time to build a typical 20-page deck.

  • Native extension advantage: Working directly in Slides eliminates export-import friction seen with Gamma, keeping edits one click away.

  • Brand guardrails baked-in: Uploading order.co’s palette, fonts, and logo lets Plus AI apply compliant styling automatically, saving manual work.

  • Flexible content control: Default mode rewrites verbose copy; strict mode preserves your words while polishing layout (ideal for board decks and other high-stakes presentations).

  • Turning memos into polished decks: Ali sometimes uses Plus AI to turn Google Docs or PDFs into first drafts of presentations, curbing blank-page paralysis.

  • Time re-allocated: A 45-minute Plus AI makeover replaced a half-day design slog, letting Ali redirect focus to higher-value strategic work.

  • Bigger picture for ops leads: Lightweight AI assistants that beautify but don’t reinvent tasks free mental bandwidth for judgment-heavy decisions across finance, customers, and strategy.

The Review

This is not a sponsored post.

What It Does

Granola is a Mac-first (but now available on Windows) “AI notepad” that sits quietly on your desktop. You jot a few quick bullets during a call, it records the audio locally (no bot joining your Zoom meeting), then expands your quick notes into a full account of the meeting, complete with action items and the ability to ask questions of the transcript.

Why Ops Leaders Should Care

  • If your life is back-to-back meetings, capturing accurate notes and action items is a must - but also a chore. Granola or another AI notetaker is starting to seem like table stakes for ambitious professionals.

  • Worried that you’re not executing quickly enough? The problem could be ineffective meetings, and an org-wide implementation of Granola (or something like it) could be the solution.

Key Features (Pros & Cons)

Pros

  • Marriage of AI and human: it records everything, but you tell it what’s important.

  • No meeting bots: Records locally, so external folks don’t get thrown off by a random guest.

  • Fast and clean: Because it’s an app that runs locally, it can run faster and with better UI than many web apps.

  • Templates + folder chat: Re-use note formats and ask the AI questions across many meetings.

Cons

  • Mac-first: if you’re a Windows user (like me), the app is still in beta.

  • Light on integrations: Slack is native, but other CRMs and project management tools will need custom Zapier connections, unlike some other tools.

  • Limited free trial: Records 25 meetings for free, then $18/month, which could be steep given free tools built into Zoom/Notion.

  • Not a fully ‘driverless’ experience: Will only nail the summary if you type a few “breadcrumb” notes (but will still be decent either way).

An Operator’s Perspective

After hearing about Granola for months, I finally gave it a shot so I could write this review. Note: I use a PC, so the Mac experience might be different. However, as a former Mac user, one of my initial impressions was that the app (which must be downloaded onto your device) feels like a Mac app - in a good way. The UI is attractive and the process is seamless - all you need to do is install it, then it notifies you about upcoming events and prompts you to start it.

The fact that it records in the background is really nice - less obtrusive and easy to forget about. The notes are quite good, but obviously don’t capture the organizational context, subtext, or body language that makes up some of the most valuable content of most meetings. The output is easy to edit and share.

I haven’t explored the enterprise-wide features, but those seem particularly valuable for capturing institutional knowledge and making it useful across companies.

Other Options

Bottom Line

If you don’t have an AI notetaker mandated by your company or haven’t fallen in love with one yet, Granola’s experience is differentiated enough from many of the other others on the market that it’s worth trying out.

The Links

News

  • After talks with OpenAI stalled, Google acquired Windsurf: The source of much consternation on the X timeline this week, the main controversy (founders and investors walking away with the bag while employees were seemingly left with zero) was at least semi-resolved when Cognition acquired the RemainCo. (Good commentary here from Kevin Kwok)

  • OpenAI set to launch a web browser: Your favorite neighborhood AI behemoth is taking another swing at Google with the pending launch of its AI-first browser. And it’s not the only one - the AI browser wars are going to heat up fast.

  •  AWS to launch agent marketplace: The partnership between Anthropic and Amazon deepens with yesterday’s launch of the agent marketplace.

  •  New Grok model launched: Grok 4 was praised for its performance but many expressed doubts about its seeming

Perspectives

  • Reflections on OpenAI: From a recently departed senior employee; Kremlinology on the AI giant is valuable for many reasons - you probably use their product, they drive the conversation in the industry, and their internal practices are likely to become widely copied.

  • Making things difficult is deliberate: In a world of ever-increasing abundance and leisure, choosing the hard path isn’t easy, but is the path to success (and happiness, probably).

Practical

  • The rise of the agent manager: ‘How many agents could you manage today?’ is a key question for you to ask as you navigate bringing AI into your org.

  • Shopify’s cultural adoption of AI: From CEO edicts to revenue teams using Cursor to hiring 1,000 interns - a lot of things to learn here.

  • How to win at context engineering: If “context is that which is scarce”, the key to building a great AI product (or even getting the most out of an LLM) is providing it with the right context. This article from Langchain goes into how you can do that.

Thanks for reading,

Tom Guthrie

If you liked this edition of AI for Operators, share it with another operator you know!