Graduation Project, 2025

VitaPsyche — AI-Powered Mental Health Platform

A graduation project combining AI diagnosis, an empathetic virtual character, and real doctor consultations to close the gap in accessible mental health care.

Next.jsReactTypeScriptDjangoMySQLMongoDBRasaHugging FaceTensorFlowFlutterFigma

Video Demonstration Showcase

Interactive Video Showcase Coming Soon

We are recording a voice-over walk-through of VitaPsyche's primary features, dashboard modules, and server operations. In the meantime, explore the complete system architecture details below!

Project Overview

VitaPsyche is a graduation project built to address the gap in mental health services by integrating AI tools, virtual support characters, and online doctor consultations into a single platform for users seeking mental health support and education. The platform combines AI-powered emotional assessment models with an empathetic virtual character named Lina for emotional support, alongside real-time consultations with psychiatrists through booking, chat, and video. As team leader, I oversaw planning, execution, and coordination across the AI, backend, frontend, and mobile tracks, using Agile methodologies coordinated through Jira and Jira Align.

Key Capabilities

AI-Powered Emotional Assessment

Built intelligent models using machine learning, NLP, and deep learning (Rasa, Hugging Face, TensorFlow) to help users assess their emotional well-being.

Virtual Support Character — Lina

Designed an empathetic AI persona that provides emotional relief and guidance to users during challenging moments.

Online Doctor Consultation

Healthcare

Enabled real-time communication with psychiatrists via appointment booking, chat, and video consultations.

Educational Resources & Self-Assessment

Content

Curated articles, FAQs, and psychological tests to help users build knowledge and self-awareness around mental health.

Interactive Architecture Flow

Hover or click on the system modules to inspect their dynamic interactions and roles within the product stack:

client

Next.js / React Web App

Web client for assessments, Lina interactions, doctor discovery, and booking.

Blueprint Sketch

Architecture mapping diagram is currently loaded in the interactive matrix viewer on the left.

Technical Challenges & Resolutions

Problem

Coordinating a cross-functional team across AI, backend, frontend, and mobile tracks toward a single cohesive product.

Resolution

Led the team using Agile methodologies with sprint planning and tracking in Jira/Jira Align, and grounded the system design in a C4 Model architecture and UML diagrams so every track could work against a shared technical blueprint.

Problem

Designing an AI persona (Lina) that feels genuinely supportive rather than a generic chatbot.

Resolution

Combined NLP and deep learning models (Rasa, Hugging Face, TensorFlow) trained toward empathetic, context-aware responses, tuned specifically for emotionally sensitive conversations.

Development timeline & Story

Phase 1: Requirements & Planning

Defining the Problem Space

Led requirements analysis and mapped the project flow using Use Case Diagrams and Process Flow Diagrams (PFD).

Phase 2: System & Architecture Design

Designing for Scale

Modeled the system with the C4 Model and UML diagrams, and designed the Django + MySQL backend for scalability and security.

Phase 3: AI Model Development

Building the Diagnosis Engine & Lina

Implemented ML/NLP/deep learning models with Rasa, Hugging Face, and TensorFlow for personalized mental health diagnosis and the Lina support character.

Phase 4: Frontend, Mobile & Integration

Bringing It Together

Built the web experience with React, Next.js, and TypeScript, and a companion Flutter mobile app, integrating booking, chat, and video consultation flows.

Interested in discussing similar system integrations?

Get In Touch All Projects