MindEase: A Local AI Mental Wellness Check-In Assistant

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle

What I Built

I built MindEase, a local AI-powered mental wellness check-in assistant that helps users reflect on their emotions through a conversational interface.

MindEase detects emotional tone and provides supportive responses, motivational quotes, calming techniques, and journaling prompts. It is designed to make everyday emotional reflection more accessible.

Key Features

• 🧠 Emotion detection for moods such as anxiety, sadness, motivation, and calm.
• 💬 Conversational memory using LangChain's ChatMessageHistory.
• 🌿 Calming suggestions, including breathing exercises, grounding techniques, and affirmations.
• ✍️ Journaling prompts and motivational messages.
• 🎨 An interactive Streamlit interface.
• 💾 Downloadable wellness logs for personal reflection.

MindEase could be useful for individuals seeking a simple wellness check-in tool or as a starting point for wellness-focused applications.

It is an experimental wellness tool, not a replacement for professional mental health care.

Demo

The application runs locally using Streamlit.

GitHub Repository: MindEase — Mental Health AI Agent

The repository includes a screenshot of the application's interface.

Code

Explore the source code and setup instructions:

https://github.com/Meghana-2124/mental_health_ai_agent

How I Built It

I built MindEase using Python, Streamlit, LangChain, TextBlob, and Ollama.

• Ollama (Gemma 2B): Runs the open-weight language model locally to generate conversational responses.
• LangChain: Supports the conversational workflow and memory management through ChatMessageHistory.
• TextBlob: Performs basic sentiment analysis.
• Streamlit: Provides the interactive user interface.
• Python: Connects the different components.

The project follows a four-layer agent architecture:

1. Input Understanding — processes the user's message.
2. State Tracker — maintains conversational context.
3. Task Planner — helps determine an appropriate response approach.
4. Output Generator — generates supportive responses using the language model.

The project also explores reflection logs, prompt iteration, and output testing.

Why Does Open Innovation Matter?

Open innovation makes AI development more accessible to individual developers who want to experiment and build practical applications without depending entirely on proprietary, paid APIs.

Using Ollama and an open-weight language model allowed me to experiment with local inference and customize the conversational workflow. LangChain and Streamlit helped me integrate the model into an interactive application.

Local inference can also reduce dependence on external AI services and help keep conversation data on the user's device, provided the application does not transmit that data elsewhere.

Building MindEase gave me an opportunity to explore how open AI technologies can be combined to create an interactive wellness application.