I like being able to pause video courses. I can replay a command, check my own terminal, or take a minute to work out what happened. But sometimes I want to ask the instructor a question before they move on. That’s what I tried using Mentor AI for while watching a course on Real Python.
Mentor AI is Real Python’s AI learning assistant, and it’s part of a Real Python membership. It knows Real Python’s content and what you’ve been learning on the site. You’ll find the Mentor AI orb in the corner of every page, and clicking it starts a conversation. You need to be signed in to use it. If you’re not a member yet, then you can start a free 7-day trial.
The course I watched was Python Project Management With uv, taught by Rodrigo Girão Serrão. Once I’d picked the course, my first question was about what Rodrigo had just said, and I ended up asking about .gitignore. The video stayed paused in the background while we talked.
You can follow along in Creating a Project With uv. Open it in a new browser tab now so you can switch between the lesson and this tutorial. If you only want to try the conversation, there’s nothing to install. You can watch Rodrigo’s terminal and read the output shown here. To code along, you’ll need uv, Git, and a terminal. Managing Python Projects With uv shows how to install uv.
The buttons in this tutorial ask about Rodrigo’s lesson. If you’re watching a different course, ask the same kinds of questions from the orb on your own lesson page instead.
This tutorial goes from picking a course to quizzing yourself after a lesson. For a list that fits where you are, ask your mentor:
Want to know what’s ahead?
Free Kit: Click here to get the Learn With Mentor AI Kit with cheat sheets, a video course, and a quiz that help you learn faster with your mentor and build the Python skills employers look for.
Note: The mentor’s replies vary from one conversation to the next. The screenshots and quotes in this tutorial come from real sessions, so the wording you get will be different.
Take the Quiz: Test your knowledge with our interactive “Learn With Mentor AI: Getting More Out of Video Courses” quiz. You’ll receive a score upon completion to help you track your learning progress:
Interactive Quiz
Learn With Mentor AI: Getting More Out of Video CoursesCheck how you use Mentor AI with Real Python video courses, from asking what the instructor said to quizzing yourself after a lesson.
Ask Which Course to Watch, and in What Order
Before you can pause a course, you have to pick one. When I wanted to learn uv, finding material wasn’t the problem, since Real Python has tutorials, video courses, quizzes, exercises, and podcast episodes about it. The hard part was knowing where to start and in which order to go through it all.
In my community video Meet Mentor AI: Your Personal Python Mentor, you can watch me ask the mentor to organize all of its uv resources by the skill level they require. I also mentioned that I like learning with video courses the most. The mentor considers your learning activity, such as the tutorials you’ve read and the quizzes you’ve taken, so the list it gave me took what I’d already done into account.
It focused on video courses, as I’d asked. But there was a nice little detail. About eleven minutes into the video, you can see it tell me that I should first finish the written tutorial, Managing Python Projects With uv, and then continue with the video course. That’s different from some other AI tools, which tend to go along with whatever you say you’d like. The mentor had a reason for the order and said so.
I took the advice and finished the tutorial before starting Rodrigo’s course. You don’t need to do the same before continuing here, since you only need the one lesson you already have open to follow along. Your own recommendation depends on what you’ve already done on Real Python, so it may look completely different. Press the button below, then edit the topic and your preferences until they’re true for you:
Most of the mentor buttons in this tutorial add to whichever conversation is open, including one you started yourself with +. Three buttons start a new conversation instead: Tell me what I’ll learn here at the top, Explain this moment in the lesson in the section about replaying the instructor, and Test my understanding at the very end.
The buttons after each of those continue in the new conversation. To get back to an earlier conversation, use the history button in the panel header.
Keep Your Mentor Out of the Way While You Watch
I don’t want a chat window taking up half the screen while I’m listening to someone, so most of the time, I keep Mentor AI minimized. You can drag its orb to the edge of the screen and park it there:

The orb slides halfway off the screen and dims, as if it’s asleep. The first time you park it, a short message such as Parked. Tap to wake me. confirms it and then fades after a few seconds. After that, the gray edge of the orb is all you’ll see, even after a reload in the same tab.
When you want the mentor back, tap the parked orb once to wake it. It slides back onto the screen, and a second click opens the chat panel. The orb’s tooltip also shows a Ctrl+M shortcut for opening the panel. After reading the answer, minimize the panel again with the arrow in its header.
You can use the mentor while watching full screen, too. The orb stays on screen, and the player’s control bar has a small orb button of its own. After a few seconds without mouse movement, the controls, the orb, and even an open chat panel fade out together. Move your pointer, and they all come back.
Typing doesn’t pause the lesson, though. Pause the video before you start typing so you don’t miss the next part while you’re asking about the last one. Or use that orb button in the control bar, which pauses the video for you. More on that button in the next section.
Ask What the Instructor Just Said
About a minute and a half into the lesson, Rodrigo lists the files that uv has just created. That’s where my first question came from: What did the instructor just say?
I didn’t type the course name or look up a timestamp. The mentor checked the page I was on and read the lesson, using my position in the video to find the passage. You can see those steps above its answer:

It picked out the explanation at 1:42. Rodrigo was showing the files uv had created, including a Git repository and a .gitignore file. The timestamp in the reply was clickable, so I could jump back and hear him explain it again. Clicking it moves the player to that timestamp and starts playback, without a page reload.
The player offers an even shorter way to ask. The orb button in its control bar, between the play button and the progress bar, is labeled Ask Mentor AI about this moment. Pressing it pauses the video, opens the chat panel, and sends I paused the video here. Can you help me with this part? for you. Here’s what that looks like in full screen at the same point in the lesson:

When I tried it, the answer summarized what happens at 1:42 and then asked what would help most, with a few suggestions to click, such as Why uv initializes Git and What belongs in .gitignore. Clicking a suggestion sends it right away.
If you have the lesson open, pause around 1:42 and try both ways: ask from the orb, and press the player’s Ask Mentor AI about this moment button. Or name the bit you missed: Why did uv create a .gitignore file? You don’t have to describe the whole video before asking about one command.
The same works when an instructor mentions a link. In the previous lesson, Installing and Updating uv, Rodrigo tells you to follow the official installation guide and says you’ll have a link handy. If you pause there and ask What was the link the instructor just mentioned?, the mentor can find it in the lesson’s resources:

Don’t rely on it for everything that comes with a lesson, though. One time, the mentor said it didn’t see any downloads for that lesson, even though the lesson’s Supporting Material menu offers the course slides and sample code. Check that menu yourself. It also has an Ask Mentor AI item of its own.
To ask about the replayed passage from here, use this button. It sends a link to Rodrigo’s lesson and the timestamp along with the question and opens a conversation of its own, so the question doesn’t land in whatever chat you had open before. That becomes your lesson conversation for the rest of this tutorial:
If you’d rather keep everything in a conversation you started on the lesson page, type the follow-up questions there instead.
Point at the Exact Sentence in the Transcript
The mentor works from the lesson’s transcript, not from the video frames, and that transcript is right there on the lesson page with a timestamp in front of every passage. If you only missed a word, the transcript is enough. But when you know exactly which sentence lost you, it’s also the most precise way to point the mentor at it. Select the passage, and an Ask Mentor AI button appears next to your selection:

The button attaches the selection to the chat and leaves the question to you. The three dots next to it open ready-made questions instead: Explain this, Break this down for me, Give me an example, Quiz me on this, What do I need to know?, and Ask something else….
If you select Rodrigo’s sentence at 1:42 and pick Explain this, the quote shows up above your message. In one conversation, the answer went a step further than the replay:

That answer explained that an ignored file stays on your computer and that .gitignore mainly keeps matching files that Git isn’t tracking yet from being added to Git. Keep that detail about untracked files in mind, because it’ll come up again.
Note: Selecting text to ask about it works on other Real Python pages, such as tutorials, too. On a lesson page, the commands Rodrigo types exist only as pixels in the video. If you want to ask about one, select it in the transcript.
Ask for a Different Explanation
A replay helps when you missed what Rodrigo said. Sometimes, though, you heard every word and still don’t understand the reason for what he’s doing.
Can you explain that another way? is a perfectly reasonable next message. So is A shorter version, please. Better still, say where you’re stuck: I can see the .git directory, but I don’t understand what it’s doing in a Python project. That question asks for more than a replay of the sentence.
If the answer starts too far back, interrupt it. Maybe you already know what version control is and only wonder about the timing. Tell the mentor what you know, and it can skip ahead:
The answer I got skipped the version control basics and went straight to the timing. uv sets Git up so the project is ready to track changes from the start, including its initial files, but “it doesn’t make a commit for you”:

A follow-up A shorter version, please. compressed that into two short sentences with the same two points.
Or ask to see it happen. An example might make more sense to you than another description of the directory:
Then go back to Rodrigo’s explanation and see whether you can follow it this time.
Chase a Side Question Without Losing Your Place
In my conversation, the replay at 1:42 gave me a reason to ask about .gitignore. It’s one of those details an instructor can mention in passing while demonstrating something else. If you haven’t used Git before, you might still be wondering what it means when the next command appears.
I asked in the same chat, and the answer appeared beside the paused lesson. The mentor went through examples of files and directories you might want Git to ignore.
Here’s the same question, Why did uv create a .gitignore file?, in a more recent conversation:

This time, the reply named the __pycache__/ folder, where Python keeps cached bytecode. It linked to Introduction to Git and GitHub for Python Developers and ended with a few follow-up suggestions to click.
Clicking How Git uses .gitignore brought back the detail from the transcript section: ignore rules apply only to untracked files, which the Git documentation on .gitignore confirms. To see the ignore rules in action, ask for a small example that shows git status before and after you add a file to .gitignore.
You can keep going with any unfamiliar term in the reply. The .gitignore that uv generates also lists .venv. Perhaps .gitignore makes sense now, but .venv doesn’t. You can follow one question through in the same conversation without starting over each time.
Note: For an unrelated side question, such as who created uv, click + in the panel header to start a new conversation. Before you press the next button in this tutorial, switch back to your lesson conversation with the history button, because buttons add to whichever conversation is open. And if an answer is worth keeping, the copy icon below it puts the text on your clipboard for your notes.
Before you go back to the video, take a minute to check yourself. Explain what you’ve learned in a few sentences of your own. I’ve prefilled an explanation with two mistakes in it. Send it as is and see what the mentor catches, or replace it with your own version:
Once you’ve seen the mentor’s reply, compare it with what a correct explanation covers. uv init doesn’t create a virtual environment. uv creates .venv the first time you run a command such as uv run, which Rodrigo covers in Running Your Project With uv. The entry is in .gitignore ahead of time, so git add skips the folder once it exists.
And adding a file that Git already tracks to .gitignore doesn’t untrack it. If you ever need to do that, ask your mentor, or see Introduction to Git and GitHub for Python Developers.
If a side question takes you further into Git than you wanted, say Keep it to what I need for this lesson. I minimized the chat panel when I was done and resumed the lesson. The Git tutorial can wait, so a question about one file doesn’t turn into an afternoon of learning Git.
Compare Your Terminal With the Recording
So far, the questions have been about what’s in the video. Coding along adds another possibility: you’ve run the command, but your terminal doesn’t look like Rodrigo’s. That doesn’t have to mean you did something wrong.
If you’re coding along, you need both uv and Git installed. Without Git, uv skips the repository and the .gitignore file. Here’s what the lesson’s command produced with uv 0.13.4, the version used for this tutorial:
$ uv init rpcats
Initialized project `rpcats` at `/home/user/projects/rpcats`
$ cd rpcats/
$ ls -A
.git .gitignore .python-version README.md pyproject.toml src
There’s no main.py. As of uv 0.12, uv init creates a packaged project by default. The code goes in src/rpcats/__init__.py, and pyproject.toml gets two extra tables:
pyproject.toml
[project.scripts]
rpcats = "rpcats:main"
[build-system]
requires = ["uv_build>=0.13.4,<0.14.0"]
build-backend = "uv_build"
The [project.scripts] entry means you’d run this project with uv run rpcats. Rodrigo recorded the course with an earlier version of uv, so his project has a top-level main.py, and the next lessons build on that file. To match the recording, create the project with the --no-package option instead. Run it in a folder that doesn’t already contain an rpcats project:
$ uv init --no-package rpcats
Initialized project `rpcats` at `/home/user/projects/rpcats`
$ ls -A rpcats/
.git .gitignore .python-version README.md main.py pyproject.toml
There are smaller differences, too. In this run, .python-version says 3.14, where Rodrigo’s says 3.13, and he points out in the lesson that yours may differ. Git may also name your branch master instead of the main in his terminal prompt, depending on your Git configuration. Only the missing main.py affects the rest of the course.
The mentor can read Rodrigo’s lesson, but it can’t inspect your local project. Asked to compare a local .gitignore with the one in the video, it once replied that it couldn’t access files on the computer and asked for the output of cat .gitignore. If the difference is in something Rodrigo shows on screen but doesn’t say, type what you see in the video, too.
To show your side, copy your uv init output together with the output of uv --version. Remove anything private, such as passwords, tokens, or personal paths. Then press the button below, paste both outputs after the prefilled text, and send:
With uv, the documentation on creating projects is worth keeping nearby. A newer release or an option you passed can produce a different project layout from the one in a recording, as the src/ package layout shows. But picking --no-package to follow the course is a trade-off:
Start by finding the reason for the difference before editing your files to match the screen. Check the mentor’s explanation against the documentation, especially if it suggests changing how your project is set up.
Quiz Yourself and Leave a Note for Next Time
When the lesson ends, a couple of minutes of recall tells you whether it actually stuck. In the course navigation below the video, next to the Overview button, another orb button opens a menu of questions about the current lesson and the whole course:

Which lesson covers…? prefills Which lesson in this course covers and waits for you to finish the sentence. Asked which lesson covers adding a command-line entry point to a project, the mentor pointed to Creating a Packageable App With uv.
For recall, use the lesson part of the menu. In one conversation, Recap this lesson returned a bulleted list of everything uv init rpcats had created, plus a link that jumped to the lesson’s summary. Quiz me on this lesson asks for one question at a time. In that quiz, the answers were multiple-choice options you could click, and the mentor gave feedback before moving on. It also pointed to the course’s Python Project Management With uv quiz:

Give a reason for each choice, even when you could just click. You’ll find out more about what you understood than you would by rereading an explanation that already sounds familiar. From here, this quiz button continues your lesson conversation, so the mentor can draw on everything you’ve asked about the lesson so far.
Think the lesson stuck?
If you’re stopping for the day, ask for a short note about the question you haven’t resolved yet. Include the lesson title and what you want to try next. On the lesson page, the mentor knows roughly where you stopped. The note I got had the lesson title, a stopping point of around 1:43, the open question, and a button to reopen the lesson:

Finish the quiz first, then ask for the note. It lands in the same conversation, so the mentor has your whole lesson chat to summarize. From this page, change the timestamp in the prefilled message to where you stopped, and finish the sentence with your own open question.
When you return a few days later, that note will be more useful than trying to remember why you stopped halfway through a video.
Conclusion
The part I appreciate most is being able to ask a short question from the lesson itself. In a separate assistant, I’d have to bring over the relevant material and explain what I was watching. Here, What did the instructor just say? was enough to start the conversation. Along the way, you used your mentor in ways that work with any video course:
- Ask which course to watch, and in what order, and take the pushback seriously when it suggests a different start.
- Keep your mentor out of the way while you watch by parking the orb, and pause before you type.
- Ask what the instructor just said from the orb or the player’s own button, and jump back with the timestamp.
- Point at the exact sentence in the transcript when you know which passage lost you.
- Ask for a different explanation that starts from what you already know.
- Chase a side question without losing your place, using + for unrelated ones, and check yourself before you resume.
- Compare your terminal with the recording by pasting your output and your tool version.
- Quiz yourself and leave a note for next time, starting from the lesson menu.
If learning in this interactive way sounds useful for you but you’re not a Real Python member yet, then start a free 7-day trial and see where learning with a personal mentor can take you.
One last check before you go. This button starts a new conversation, so the mentor checks what you can explain without leaning on your earlier answers:
Ready to prove it?
Next time an instructor mentions a term you don’t recognize, pause there and give it a try. And when you practice what you’ve learned, your mentor can also help you work through coding exercises with hints instead of answers.
Free Learn With Mentor AI Kit: Video, Cheat Sheets, Course, and Quiz
Learn Python faster by working through it in conversation with Mentor AI, and focus your practice on the Python skills employers look for.
Take the Quiz: Test your knowledge with our interactive “Learn With Mentor AI: Getting More Out of Video Courses” quiz. You’ll receive a score upon completion to help you track your learning progress:
Interactive Quiz
Learn With Mentor AI: Getting More Out of Video CoursesCheck how you use Mentor AI with Real Python video courses, from asking what the instructor said to quizzing yourself after a lesson.