🎙️🧠 CBT Voice Agent

CBT Voice Agent is a voice-enabled assistant that provides personalized support using Cognitive Behavioral Therapy (CBT). It allows users to express their feelings verbally, processes the input using an agentic workflow powered by the Deepseek-R1 LLM and LangGraph library, and delivers supportive, actionable advice through audio responses.

---

Project Demo:

Click on the image to watch the project demo on YouTube:

---

🚀 Features

• 🎤 Voice Interaction:

  • Speak directly to the app to express your feelings or concerns.
  • 📝 Speech is transcribed to text using the OpenAI Whisper model.

• 🤖 AI-driven CBT Support:

  • Uses a structured, agentic workflow to provide targeted CBT interventions.

• 🗣️ Natural Conversational Responses:

  • Delivers empathetic and human-like audio replies.
  • 🔊 The ElevenLabs library is used for text-to-speech.

• 📜 Chat History:

  • Remembers the context of previous interactions.

---

⚙️ Understanding the Agentic Workflow:

The workflow is created using the LangGraph library. It uses three specialized AI agents. Each agent is powered by the Deepseek-R1 LLM. They work in this sequence:

1️⃣ Analysis Agent 🔍

• Functionality:

  • Processes the transcribed user query.
  • Identifies underlying emotions, cognitive distortions, and thought patterns.

• Output:
  Provides a structured analysis of the user's emotional state.

2️⃣ Intervention Agent 🛠️

• Functionality:

  • Receives analysis from the Analysis Agent.
  • Designs targeted CBT interventions (e.g., cognitive restructuring, behavioral activation).
  • Uses structured Socratic questioning or CBT activities to guide the user.

• Output:
  Generates a clear, actionable CBT strategy for the user's specific issue.

3️⃣ Support & Feedback Agent 💬

• Functionality:

  • Receives the intervention from the Intervention Agent.
  • Adds empathetic validation, encouragement, and supportive feedback.
  • Prompts the user for reflection on their emotional state and intervention effectiveness.

• Output:
  Delivers a supportive, empathetic, and complete audio response to the user.

---

🗃️ Agentic Workflow Structure

User Input (Audio) 🎙️
      │
      ├──► Speech-to-Text 🎧
      │
      └──► Analysis Agent 🔍
                  │
                  └──► Intervention Agent 🛠️
                              │
                              └──► Support & Feedback Agent 💬
                                          │
                                          ├──► Text-to-Speech 🎧
                                          │
                                          └──► User Receives Audio Response 🎙️

---

🧠 Why CBT?

Cognitive Behavioral Therapy (CBT) helps people change negative thought patterns and develop healthier ways of thinking. This app uses CBT to:

• 🧐 Recognize Negative Thinking: Identifies unhelpful thoughts that may be affecting your emotions.
• 🔄 Challenge and Replace Them: Helps you question if those thoughts are true and guides you to a more balanced perspective.
• 🏆 Provide Practical Coping Strategies: Suggests real actions you can take to improve your mood and manage stress.
• 🤗 Offer Encouragement and Support: Gives positive feedback and motivation to help you feel better.

By following CBT techniques, the app helps you think more healthily and take steps toward feeling more in control.

---

📁 Project Structure

📂 CBT_Voice_Agent
│
├── 📄 LICENSE              # License information
├── 📄 README.md            # Project overview and documentation
├── 📄 requirements.txt     # Python dependencies
├── 📄 cbt_voice_agent.py   # Main application script (Streamlit app & agentic workflow)
└── 📄 prompts.py           # Prompt definitions for Analysis, Intervention, and Support Agents

---

📜 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

---

Contact

For questions or suggestions, please open an issue or contact me directly:

• Email: adnanbarwaniwala7@gmail.com

---

🙏 Thank You

Thank you for spending time on my repo. Hope you enjoyed it!