WithMe AI

Inspiration

Mental health challenges affect millions of people every day, yet many people struggle to find someone to talk to when they need support. While AI assistants like ChatGPT and Gemini can hold conversations, they are designed for general-purpose tasks. We wanted to create something focused specifically on emotional well-being.

WithMe AI was inspired by a simple question:

What if people had a safe, supportive space to reflect on their thoughts before they become overwhelmed?

Our goal was not to replace professional help, but to create a companion that encourages self-reflection, emotional awareness, and small positive actions.

What it does

WithMe AI is an AI-powered emotional support companion.

Users can:

Talk about their thoughts, worries, and daily challenges
Receive empathetic and reflective responses
Explore their emotions through guided reflection
Identify factors contributing to how they feel
Focus on small, achievable next steps

The platform is designed to listen first, reflect second, and advise only when appropriate.

How we built it

We built WithMe AI using:

HTML
CSS
JavaScript
Python
Flask
Gemini AI
GitHub
Render

The frontend was designed with a calm and minimal interface to create a comfortable environment for users.

The backend uses Flask and Gemini AI to generate supportive responses while maintaining a structured reflection process.

The application was deployed using Render to provide a live, accessible experience.

Challenges we ran into

One of the biggest challenges was designing an AI that felt supportive without sounding robotic or overly clinical.

Early versions often produced generic responses that felt similar to traditional chatbots. We spent significant time refining prompts and conversation flows so that responses focused on understanding the emotions behind a user's situation rather than simply reacting to keywords.

Deployment was another challenge. Moving from frontend-only projects to a full-stack application required learning backend deployment and hosting with Render.

We also spent considerable time improving the visual design to create an experience that felt calm, welcoming, and trustworthy.

Accomplishments that we're proud of

Building a complete AI-powered mental wellness platform
Creating a reflection system that goes beyond simple chatbot conversations
Successfully deploying a Flask backend using Render
Designing an interface focused on comfort and accessibility
Developing a project that encourages emotional awareness and self-reflection

What we learned

Through this project we learned:

Full-stack web development
Backend deployment with Render
Prompt engineering and AI behavior design
User experience design for mental wellness applications
The importance of responsible AI and thoughtful conversation design

We also learned that creating a good mental health experience is not only about technology. It is about understanding people, emotions, and how users interact with supportive tools.