MindBridge

In Morocco, accessing mental health care is often difficult. Patients may wait weeks or even months between sessions, and when they finally meet a doctor, the time is limited to less than an hour. We realized a critical gap: mental health happens every day, but care happens only occasionally. Important emotional patterns, stress triggers, and daily struggles are often lost between sessions. This inspired us to build MindBridge a system that bridges this gap and helps doctors better understand their patients beyond the limited session time.

AI Architecture (3 Agents)

Under the hood, MindBridge is a multi-agent system designed for clinical use, with built-in safety checks. We use three specialized AI agents working together:

    **Decomposer Agent**

-Takes the doctor's complex question.
-Breaks it into 3 to 5 clear, structured goals.
-Ensures consistency and avoids random or unclear outputs.

    **Conversation Agent**

-Interacts with the patient one goal at a time.
-Uses natural, supportive, and emotionally aware conversations.
-Acts as a guided interviewer, not just a chatbot.
-Leverages patient history and context for better understanding.
-Includes safety and quality checks on every response.
-Detects emotional distress and potential risk signals.

  ** Summarizer Agent**

-Collects all insights from conversations.
-Transforms them into a clear, structured report.
-Provides direct, useful answers to the doctor's original question.

MindBridge turns scattered emotions into clear clinical insight—so every session starts with understanding, not guesswork.

Built With

• firebase
• gemini-api
• llama-3.3-70b-versatile-api
• react.js
• tailwind