How to Seamlessly Record Conversations with Chatbots and Camera Viewer

Introduction

I’ve been searching for this solution for over a year. The problem: I wanted to record my AI assistant farmhand (pi.ai) while I was video blogging my farm chores, and use that content for my YouTube videos. If you’ve ever wanted to record conversations with chat tools like Pi.ai, ChatGPT, or other virtual companions while also displaying your camera in real-time, you know how tricky it can be. Managing multiple apps and ensuring everything stays active without interruptions is a real challenge, especially on mobile devices like **Samsung phones**.

The problem is, Samsung and many other phones doesn’t allow the camera mic to hear the apps and websites. It’s one or the other. I knew it was possible because the screen recorder could offer me a tiny cam view overlay while chatting with Pi, but I tried every screen recorder app and every camera app and every Bluetooth option known to the world, Reddit, Discord and even Chat GPT. It seemed an impossible mission.

I saw dozens of “me too” and “Did you ever find a solution posts, but nobody could get the camera and Pi.ai working. Pi didn’t want to run in the background either. Chat GPT now does voice and runs in the background but only recently.

I don’t give up on projects like this so every few weeks I’d start over and see what else I could try. Then one day I woke up and tried something new. If a solution didn’t seem to exist, maybe I could create it. I’m not a programmer, but I do spend a lot of time learning about new technology and AI tools.

I visited https://websim.ai and simply asked, “Can you create a web page that displays my camera but does not record sound?” In three seconds, I had a web page. After two revisions, I had a silent camera app that worked on Windows and probably any phone. On my Samsung, it was a one-click silent camera that let me select all four cameras by hand. Not even the Samsung camera lets me choose a wide angle for selfies.

Now I had a working solution that was “nearly” perfect. I still can’t get Pi.ai to run in the background, but my camera opens in a preset split screen mode with a small window at the bottom for Pi to stay active. Next I created a routine on my Samsung so that a single click loads Pi, the camera and starts recording sound and video and Pi. I can flip to any camera smoothly and when Pi stalls, I can refresh her without having to task flip off the camera. When I quit, it all shuts down.

I click one button and it’s all ready to go. I can record videos with the cows and chickens while Pi tells me about Angus cattle histories and what foods were bad for our horse. If you’ve ever wanted to record conversations with chat tools like Pi.ai, ChatGPT, or other virtual companions while also displaying your camera in real-time, you know how tricky it can be. Managing multiple apps and ensuring everything stays active without interruptions is a real challenge, especially on mobile devices like **Samsung phones**.

Websim.ai is an awesome tool. Fun and useful and easy. It is addictive. I also used ChatGPT to brainstorm because this kind of work needs a second person to work things out and for me, AI chat companions are equally suitable if not better.

### **What This Setup Achieves**
This approach is designed to:
– **Launch all necessary apps** (screen recorder, chatbot, and camera viewer) in sequence.
– **Keep the chatbot active and accessible** in split-screen mode or pop-up view.
– **Record conversations easily** with one-tap automation.
– **Capture the camera feed with full-screen flexibility**, even allowing multiple cameras to display—a feature not natively available in many camera apps.

• I did not try to get the dual camera mode working. Maybe tomorrow.

### **What You’ll Need**

Although I designed this to suit my specific needs, it also should work on any phone, Android or iPhones, Mac and Windows. If you don’t chat with Pi.ai, it will work with any program that uses a mic and audio with your camera. The routines portion that automates the opening process is just a nice Samsung bonus but on other phones, you just open the web page, screen recorder and chatbot software manually.

– A **Samsung phone** with **Modes and Routines** enabled.
– **Pi.ai** (or any other chatbot) installed and ready to use.
– **Samsung Internet** (or another browser) with the camera viewer set as the homepage.
Because it is just a web page, I used a different browser so that I could save the split screen locations for this app and no others.

– **Screen Recorder** (pre-installed on Samsung phones).

### **Step-by-Step Setup**

#### **1. Prepare the Camera Viewer**

1. Open the camera viewer URL in your browser. It is a single HTML file you can load locally or add to any web page. You can use mine: https://frogstar,tv/cam.html

2. **Set it as the homepage if you plan to automate the routine, or save it as a desktop icon or bookmark.

#### **2. Create a Samsung Routine**
Samsung’s **Modes and Routines** feature allows you to launch multiple apps in sequence with a single tap. Here’s how to set it up:

1. **Open Samsung Routines**.
2. **Tap the + button** to create a new routine.
3. **Trigger**: Set **”When I tap a button”** or a **custom trigger**.
4. **Add Actions**:
– Launch **Screen Recorder**.
– Open **Pi.ai**.
– Open **Samsung Internet** (which will load the camera viewer automatically).
5. **Save the Routine** with a name like “Chat & Record.”

—

—

### **Bonus: Use App Pairs for Faster Access**
If you use **split-screen frequently**, save the layout as an **App Pair**:
– Open **Pi.ai** and the camera viewer in split-screen.
– **Tap the divider** between the two apps and choose **Create App Pair**.
– This will create a shortcut on your **home screen** to instantly launch both apps in split-screen mode.

—

### **How It All Works in Practice**
Now, with just **one tap on the routine button**, everything opens in sequence:
1. **Screen Recorder** starts.
2. **Pi.ai** launches in **split-screen** or **pop-up view**.
3. **Samsung Internet** loads the **camera viewer**, which supports multiple cameras.

This streamlined approach lets you chat with Pi.ai (or any chatbot) while **recording video from your camera feed**—and it’s all done without any interruptions. If Pi.ai crashes (which can happen with poor internet), the split-screen makes it easy to **tap back in and restart quickly** without missing a beat.

—

### **Conclusion**
This setup has solved multiple challenges for me, making it easy to **chat, record, and capture my camera feed all in one go**. It works well not only with Pi.ai but also with **ChatGPT** or even virtual companions, providing a flexible tool for anyone wanting to record conversations.

If you’re struggling with **keeping chatbots active** while recording on your Samsung device, give this workflow a try! It’s a great way to **simplify interactions**, stay productive, and capture moments that matter—whether for documentation, reflection, or just for fun.

—

I hope this post helps others find a **simple, practical solution** for their chatbot recording needs. If you have any questions or need more details, feel free to reach out!

—

Let me know if you’d like to tweak any part of this post before publishing, or if there’s anything else you’d like to add!

Note: on Windows I use this as well, and if you  have more than one camera, you can open as many versions of the cam.html as you want. You can chat with Animaze versions of Pi, or Adone Character animator versions of Pi or Xpression Camera versions, Snapcan versions, Xsplit versions… you get the picture.