🚀 AskGPT: Your AI-Powered Live Call Assistant 🎙️💬

AskGPT is a real-time AI assistant that enhances your calls, interviews, meetings, and lectures by providing instant AI-generated insights.
It captures live audio, transcribes speech into text, and generates context-aware responses—all in real-time!

✨ Features

• 🎤 Live Transcription – Converts speech into text instantly using Whisper AI.
• 🤖 Instant AI Responses – Get context-aware AI-generated answers on the fly.
• ⚙️ Customizable AI Context – Adapt AskGPT for customer service, interviews, or important calls.
• 🔒 Local & Private – Runs entirely on your machine using Ollama, keeping data secure.
• ⚡ Fast & Seamless – Streams AI responses token-by-token for real-time assistance.

---

🔍 How It Works

1. Start AskGPT – It listens to your call or meeting.
2. Transcribes Speech – Converts spoken words into editable text.
3. Modify & Ask AI – Adjust the text before sending it to AI.
4. Instant AI Assistance – AI generates short, precise answers with brief explanations.

💡 Use it for:
✅ Providing Context for Customer Service – Quickly retrieve relevant information to assist customers efficiently.
✅ Getting Context on Important Calls – Summarize key points and remember critical details.
✅ Help During Interviews – Get AI assistance in answering behavioral, technical, or case study questions.

---

🎧 Setting Up BlackHole on macOS for Audio Routing

AskGPT uses BlackHole to capture system audio for transcription. Follow these steps to install and configure it properly.

---

1️⃣ Install BlackHole

BlackHole is a virtual audio driver that allows applications to send and receive audio between different apps.

1. Download BlackHole from the official BlackHole GitHub.
2. Install it by following the instructions in the .pkg installer.

---

2️⃣ Create a Multi-Output Device (Route Audio to Both Speakers & BlackHole)

To hear the system audio while also sending it to BlackHole:

1. Open Audio MIDI Setup (Command + Space → search for "Audio MIDI Setup").
2. Click + in the bottom-left corner and select "Create Multi-Output Device".
3. In the right panel, enable:
   • ✅ Built-in Output (Your Mac’s speakers or external device)
   • ✅ BlackHole 16ch (or the version you installed)
4. Set BlackHole as the Master Device.
5. Check Drift Correction for your built-in speakers.
6. Rename it (optional) to something like "BlackHole + Speakers" for easy identification.

---

3️⃣ Find & Set BLACKHOLE_INDEX in app.py

AskGPT needs the correct input device index to listen to BlackHole.

✅ Find the BlackHole Device Index

Run the following script to list available audio devices:

python devs.py

This will output something like:

Index 0: Built-in Microphone
Index 1: BlackHole 16ch
Index 2: Built-in Output

Look for the BlackHole device and note its index number.

✅ Update AUDIO_DEVICE_INDEX in app.py

Open app.py and update this line:

AUDIO_DEVICE_INDEX = X  # Change this to your BlackHole device index

For example, if BlackHole 16ch is at index 1, update:

AUDIO_DEVICE_INDEX = 1  # Change this to your BlackHole device index

Save the file and restart AskGPT.

---

4️⃣ Set Your Mac’s Audio Output to the Multi-Output Device

1. Open System Settings → Sound.
2. Under Output, select your Multi-Output Device.
3. Now, all system audio will be played through both your speakers and BlackHole.

---

5️⃣ Use BlackHole as an Input in AskGPT

1. Open AskGPT and ensure it is set to listen to BlackHole.
2. Now, the system audio will be transcribed while still being audible through your speakers.

💪 You’re all set! Now, AskGPT will capture system audio while you can still hear it! 🎧🔥

---

🛠️ Troubleshooting

• If you don’t hear sound, check that Built-in Output is enabled in your Multi-Output Device.
• If AskGPT doesn’t capture audio, ensure BlackHole is selected as the input device in your app.
• Run devs.py again if BlackHole's index changes after reconnecting devices.
• Restart your Mac if changes don’t apply immediately.

🛠 Installation & Setup

1️⃣ Install Dependencies

Make sure you have Python 3.9+ installed. Then, install required libraries:

pip install flask flask-socketio pyaudio whisper flask-cors requests

2️⃣ Install & Run Ollama

Download Ollama and install the best AI model for your use case:

curl -fsSL https://ollama.ai/install.sh | sh

Then, install the AI models:

ollama pull codellama  # Best for coding-related queries
ollama pull mistral    # Fast & balanced responses
ollama pull gemma      # General-purpose AI model

Now, start the Ollama server:

ollama serve

3️⃣ Run AskGPT

python app.py

Then open:

http://127.0.0.1:8080

---

⚙️ Customizing the AI Context

Click "Set Context" in the UI to modify the AI’s behavior. You can:

• 🏆 Support customer service interactions (e.g., "Summarize key customer concerns and suggest appropriate responses.")
• 📝 Assist in important calls (e.g., "Extract important details and action points from business meetings.")
• 🎤 Help during interviews (e.g., "Provide structured responses to behavioral and technical interview questions.")

---

🎯 Who Should Use AskGPT?

🚀 Customer Service Representatives – Get real-time AI-generated insights to improve responses.
📞 Professionals & Business Users – Take live notes and gain context during important calls.
🎙️ Job Seekers & Interviewees – Receive AI-assisted responses in interviews beyond technical roles.

No more struggling to keep up—AskGPT is your personal AI assistant for live conversations! 🚀🔥

---

🤝 Contributing

Got an idea for improvement?
Feel free to submit a pull request or open an issue on GitHub!

---

📜 License

CC0 1.0 Universal – Free for personal and commercial use.