Back to Projects
Artificial Intelligence
Completed

MindHaven – AI Mental Health Companion

An AI-powered mental health chatbot that provides empathetic, context-aware conversations using LLaMA 3.2, LangChain, and sentiment analysis.

March 2026• Team: Personal Project• Role: AI & Python Developer
MindHaven – AI Mental Health Companion

About this Project

MindHaven is an AI-powered desktop application designed to provide users with a safe and supportive space to express their thoughts and emotions. Powered by LLaMA 3.2 through Ollama and enhanced with LangChain's conversational memory, the chatbot delivers intelligent, context-aware interactions. TextBlob-based sentiment analysis allows the assistant to understand the user's emotional state and generate more empathetic responses, while a modern Tkinter interface ensures an intuitive and privacy-focused user experience.

Tech Stack

Python
LangChain
LLaMA 3.2
Ollama
TextBlob
Tkinter
Pillow

Tools Used

VS Code
Git
Ollama
Python

Key Features

Core Features

  • ▸AI Mental Health Companion: Developed 'Man-Sathi', an intelligent chatbot that provides empathetic and context-aware conversations for emotional well-being.
  • ▸Emotion-Aware Responses: Integrated TextBlob sentiment analysis to understand user emotions and adapt the chatbot's response tone.
  • ▸Contextual Conversations: Utilized LangChain memory with LLaMA 3.2 to maintain natural multi-turn conversations and improve user experience.

Architecture & Technology

  • ▸Local AI Processing: Powered by Ollama to run the LLaMA 3.2 model entirely on-device, ensuring complete user privacy.
  • ▸Desktop Interface: Designed a clean dark-themed GUI using Tkinter and Pillow for an intuitive chat experience.
  • ▸Modular Design: Separated AI logic, sentiment analysis, and user interface into independent modules for scalability and maintainability.

Learning Outcomes

  • ▸Large Language Models: Learned practical integration of LLMs using Ollama and LangChain.
  • ▸Natural Language Processing: Applied sentiment analysis to build emotionally intelligent conversational systems.
  • ▸Desktop Application Development: Strengthened Python GUI development and backend integration skills.

Highlights

AI Chatbot
Sentiment Analysis
Local LLM
Desktop Application

Installation

Clone the Repository

git clone <YOUR_GITHUB_REPOSITORY>
cd mindhaven

Install Dependencies

pip install -r requirements.txt

Start Ollama

ollama run llama3.2

Launch the Application

python main.py

Challenges & Solutions

Challenge

Maintaining natural conversations across multiple user interactions

Solution

Integrated LangChain's conversational memory to preserve context, allowing the chatbot to generate coherent and personalized responses.

Challenge

Understanding users' emotional state during conversations

Solution

Implemented TextBlob sentiment analysis to detect emotional tone and subtly influence the chatbot's responses with greater empathy.

Challenge

Ensuring complete user privacy while using AI models

Solution

Leveraged Ollama to run the LLaMA 3.2 model locally, eliminating the need for cloud-based AI services and keeping all conversations on the user's device.

Challenge

Building an intuitive desktop interface for AI interactions

Solution

Designed a responsive Tkinter-based GUI with a modern dark theme, making conversations simple, accessible, and user-friendly.

LinkedIn
GitHub