---
title: 'Perplexity Just Took AI Automation to the NEXT LEVEL!'
source: 'https://youtube.com/watch?v=bXxFiETolyE'
video_id: 'bXxFiETolyE'
date: 2026-11-17
duration_sec: 564
channel: 'Julian Goldie SEO'
---

# Perplexity Just Took AI Automation to the NEXT LEVEL!

> Source: [Perplexity Just Took AI Automation to the NEXT LEVEL!](https://youtube.com/watch?v=bXxFiETolyE)

## Summary

Perplexity has introduced 'Teammates' in its Mac app, a feature that shifts AI interaction from one-off prompts to ongoing, goal-oriented jobs. Teammates work autonomously in the background, can collaborate with other agents, and check in with users for approval on key decisions. Currently exclusive to Max subscribers, the feature represents a significant step toward AI agents that handle continuous work.

### Key Points

- **Teammates arrive in the Mac app** [00:02] — Perplexity announced on November 13th that Teammates are now available in the Perplexity Mac app for Mac subscribers. The feature first appeared on the web earlier in the month.
- **What is a Teammate?** [00:58] — Unlike a chat prompt, a Teammate is given an ongoing job (e.g., QA, customer service, research, growth marketing). You provide a high-level goal; the teammate plans the work, runs tasks side by side, and can add more agents if needed.
- **Shift from prompts to goals** [01:44] — Perplexity CEO Aravind Srinivas described Teammates as 'personal agents for work,' changing the interaction from writing prompts in a session to giving high-level goals. This reduces the attention required for step-by-step prompting.
- **How Teammates work** [02:12] — Each teammate maintains a persistent thread, works in the background on its own cloud computer, has a separate desktop and browser, and can use existing projects, skills, memory, and tools.
- **What's new vs. Perplexity Computer** [03:48] — Teammates add multi-agent teamwork, full desktop and browser use, and proactive monitoring. They can also message co-workers in Slack and work from memory and prior context.
- **Approval before action** [04:28] — Teammates ask for approval before acting on decisions, keeping the user in the loop. You review and approve before it moves forward.
- **Example jobs** [04:45] — Perplexity's examples include an on-call engineer that handles alerts, a QA engineer that tests fixes, a legal contract reviewer, an email/calendar checker, and a competitor tracker. These illustrate the 'role, job, check-in' shape.
- **Availability and limitations** [06:07] — Teammates in the Mac app are currently exclusive to Max subscribers. Perplexity plans to widen access but hasn't shared a date for Pro. The number of teammates per account is undisclosed, and the feature is a research preview.
- **How to get ready** [07:03] — Three steps: practice writing goals instead of prompts, pick one repeating job and define it clearly, and decide where you want human check-in points. These skills apply to any agent tool.

### Conclusion

Teammates represent a shift toward AI agents that own ongoing work rather than single answers. Even if you're not on Max, the habit of giving clear goals and reviewing results is a skill you can start building today.

## Transcript

the next level. Ever wished you could give an AI a job title, not a prompt. give an AI a job title, not a prompt. QA, customer service, research, growth marketing. Perplexity just made that real and it's now sitting inside the Mac
app. But there's a gate on the door. Today, I'm breaking down teammates inside the Perplexity Mac app. I'll show you what they actually are, how they keep working when you're not even there, and the one rule that decides who gets
to use them. And near the end, I'll give you three simple things you can do right now to get ready. Stay with me. Okay, here's what happened. On November 13th, Perplexity posted on X. Teammates are now available in the Perplexity Mac app
for Mac subscribers. Short post, big idea. Teammates first showed up on the web earlier this month. Now they're on the Mac, too. So, what is a teammate? Think of it like this. Most AI tools work in a chat. You type a prompt, you
get an answer, then you type again. A teammate is different. You give it an ongoing job, not a one-off question, a job. Perplexity says teammates take on work for you. They can work on their own or they can team up with other
teammates. And Perplexity names jobs like QA, customer service, research, and growth marketing. Here's the part I like. You don't guide it step by step. You give it a highle goal. The teammate plans the work. It runs tasks side by
side. And it can add more agents if the job gets bigger. Then it keeps going until the results are ready for you to review. On launch day, Perplexity CEO Aravvinas called teammates personal agents for work. He said they changed
things from writing prompts in a session to giving highlevel goals. So, the shift is simple. Before you wrote prompts inside a session, now you hand over a goal. And that matters because prompting eats your attention every single step. A
teammate is built to take some of that off your plate. One more piece of the picture. Teammates work inside Perplexity Computer, the tool Perplexity already has for carrying out tasks. That's the big picture. Let's get into
how a teammate actually works. First, each teammate keeps one thread that never closes. Think of it like a running chat with a co-orker. You can ask what it's doing, what it finished, and when. You can also change its plans anytime.
Second, it works in the background. Digital Trends reports that each agent runs on its own cloud computer, so the work keeps going even when your laptop is closed. That's the big difference from a normal chat window. Third, each
teammate gets its own desktop and its own browser. Perplexity says browsing sessions and cookies are kept separate for each one so they don't tangle with each other. Fourth, it can use what you already built. Perplexity says each
teammate draws on your computer tasks, skills, memory, and tools. Projects, skills, and memory carry over, and teammates use the skills and plugins you install with the same permissions. Quick pause before I show you the next part.
If you want to learn tools like this properly, we've got a 30-day road map inside the AI profit boardroom. It takes you through learning AI agent tools one step at a time, so you're never staring at a blank screen wondering where to
start. There are also four coaching calls every week, so when you get stuck on a new tool like Teammates, you can bring your questions to a live call and get unstuck. It ties right into what you just learned. Agents work best when you
know how to direct them. link in the description or go to aiprofitboardroom.com. All right, back to it. Now, if you've used Perplexity Computer, you might ask, "What's new?" Perplexity says teammates
add three things: multi-agent teamwork, full desktop and browser use, and proactive monitoring, so a teammate can keep watch on something instead of waiting for you to ask. Fifth, they can talk to people.
Teammates can work with other teammates and they can message co-workers in and they can message co-workers in Slack. Sixth, memory. Perplexity says teammates work from memory and prior context. So, you're not explaining
everything from scratch each time. It also says teammates can pick up projects or recurring tasks that others started. And here's the detail I think matters most. Teammates ask for approval before acting on decisions. So, you stay in the
loop. You review, you approve, then it moves. So, what jobs does Perplexity stick to what Perplexity itself has published. On its teammates page, Perplexity shows an on call engineer. When an alert fires, it gathers logs,
finds the root cause, and opens a pull request with a fix and a plan to roll it back. Perplexity says it only contacts you if the fix needs approval or if it can't find the cause within 15 minutes. There's also a QA engineer that tests
There's also a QA engineer that tests fixes on desktop and mobile and a legal contract against your standard terms and sends back a red line. In its update notes, Perplexity lists two everyday jobs. One checks your email and calendar
each morning, flags messages that need replies with drafts ready, and preps notes for your meetings. The other tracks named competitors for launches and new hires, alerts you right away, and sends a weekly summary. Now, to be
clear, these are Perplexity's own examples. I'm not claiming results from them. They just show the shape of the idea, a role, a clear job, a point where it checks in with you. Why does that shape matter? Because a lot of work
isn't hard. It's just constant. Watching an inbox, keeping an eye on a list of competitors, running the same checks again and again. Teammates are built for that kind of ongoing work, so more of your attention can go to the parts that
actually need you. Now, let's talk about the gate on the door. Here's the timeline. On November 5th, Perplexity CEO said teammates were rolling out to Max users on the web and that native apps for mobile and desktop were coming
soon. 8 days later, the Mac app got them. Right now, teammates in the Mac app are for Mac subscribers only. That's Perplexity's top plan. If you're on a lower plan, you don't get in yet. One report says Perplexity plans to widen
access in the coming weeks, but it hasn't shared a date for Pro. Here's another honest note. I couldn't find how many teammates One Max account can run. Perplexity hasn't shared that number in the places I checked. That same report
also calls this a research preview, so expect things to change. Buttons move, features shift. That's normal for something this new. And one more thing, a teammate can do a lot on its own. That's exactly why the approval step
matters. Read what it plans to do before you say yes. So, is this worth learning about even if you're not on max? I think yes, because the idea behind it isn't going away. So, how do you get ready even if you can't use teammates today?
Number one, practice writing goals instead of prompts. A prompt asks for one answer. A goal describes the outcome you want over time. You're describing the result, not every step. Try rewriting a few of your own prompts that
way. Number two, pick one job that repeats, something you do every day or every week. Write it down. What starts it? What does done look like? What tools does it touch? A teammate needs a clear job, and so do you before you hand it
over. Number three, decide where you want a human in the loop. Perplexity's examples build in check-in points, an approval, a review, a message when something looks off. Plan yours first. Then it's much easier to trust whatever
you build. These skills work with any agent tool, not only this one. Perplexity is clearly moving toward agents that own ongoing work, not single answers. If you learn to hand off a job clearly, you're ready whichever tool you
end up using. So, here's the bottom line. Teammates in the Mac app are max only for now, and they're still new. But the habit they point to, giving goals and reviewing results, is one you can start building today. Here's the thing.
business owners haven't figured this out yet. And that window won't stay open forever. This is about getting ahead on learning how this kind of tool works. No promises on results, just a head start. Inside the AI profit boardroom, you get
daily tutorials that walk you through AI tools, four coaching calls every week where you bring your questions and get unstuck, a prompt library so you're not starting from a blank page, and the member map, a way to find and connect
description or go to aiprofitboardroom.com. If you want the full process, the SOPs, and over 100 AI use cases like this one, join the AI success lab. It's our free community. Links in the comments and
description. You'll get all the video notes from this episode, plus access to 95,200 members. The AI Success Lab is free to join. The video notes put everything from this episode in one place, so you
can come back to the details whenever you need them. If you're learning AI tools, it's a simple place to start alongside other people doing the same. So, grab the link in the comments or description, and I'll see you
