amber-system

Nikhath MF

Build a web-based system called **MindCare AI**, which performs mental health risk assessment and provides basic guidance to users. The goal is not medical diagnosis but awareness and early intervention. Start by designing the system as a structured pipeline. The user should first interact with a questionnaire-based interface. This questionnaire must follow a standardized scale (0–3 rating per question) and should cover key indicators such as mood, sleep, energy, and concentration. The input must be clean, validated, and consistent. Once the user submits responses, the backend should process the data in two ways. First, implement a rule-based scoring system that calculates the total score and maps it into risk categories such as low, moderate, or high. Second, integrate a machine learning model (preferably SVM or Logistic Regression) that takes the same inputs and predicts a risk level. Combine both outputs using a safe decision rule (for example, selecting the higher severity level) to avoid underestimating risk. After determining the final risk category, pass the result to a recommendation engine. This component should generate structured output based on severity. For low risk, provide general wellness tips like maintaining routine and physical activity. For moderate risk, include coping strategies such as journaling, meditation, or talking to trusted individuals. For high risk, the system must strongly recommend seeking professional help and display a clear warning message. In parallel, implement a mood tracking feature. Users should be able to log their daily emotional state. Store this data in a lightweight database such as SQLite. This data will later be used to build a dashboard where users can view their mood history over time. If possible, include a simple visualization like a trend graph to enhance usability and presentation quality. From a frontend perspective, the application must include the following pages: a landing page with a clear purpose and navigation, an assessment page with structured questions, a results page showing risk level and recommendations, and a dashboard page displaying stored mood data. The UI should be clean, minimal, and consistent, using calm colors and readable typography. There are several mandatory requirements that must not be skipped. The system must include a disclaimer clearly stating that it is not a medical diagnostic tool. Input validation must be enforced to prevent invalid or missing data. The system should avoid collecting personally identifiable information to maintain basic privacy standards. Additionally, the logic must always prioritize user safety by not underestimating risk levels. The final system should demonstrate a hybrid approach combining rule-based logic and machine learning, persistent data handling through a database, and a user-friendly interface that supports both assessment and tracking. The focus should be on clarity, usability, and responsible output rather than claiming high medical accuracy. The end result should be a fully functional website that allows users to assess their mental health risk, receive actionable suggestions, and track their emotional patterns over time.

Comments (0)

No comments yet. Be the first!

Page 1 of 6

System Requirements Document (SRD)

Project Name: amber-system

1. Introduction

The amber-system is a web-based platform designed to perform mental health risk assessments and provide basic guidance to users. The system aims to raise awareness and encourage early intervention for mental health concerns. It is not a diagnostic tool but rather an educational and supportive resource.

The system will feature a structured hybrid pipeline combining rule-based logic and machine learning to assess mental health risks. Users will also have access to a mood tracking dashboard to monitor their emotional patterns over time. The website will prioritize user safety, privacy, and usability, ensuring a seamless and responsible experience.

Page 2 of 6

2. System Overview

The amber-system is designed to:

Provide a mental health risk assessment through a 10-question standardized questionnaire.
Process user responses using a hybrid pipeline that combines rule-based scoring and a Logistic Regression machine learning model.
Deliver tailored recommendations based on the assessed risk level, with a clear warning for high-risk users.
Allow users to log their daily mood anonymously and view trends over time via a dashboard.
Maintain user privacy by avoiding the collection of personally identifiable information (PII).

The website will be built using modern web technologies, including FastAPI for the backend, React for the frontend, and MongoDB for data storage. The design will follow a "Midnight Zen" dark theme, ensuring a calming and user-friendly interface.

3. Functional Requirements

As a User, I should be able to complete a 10-question mental health assessment with a 0–3 rating scale.
As a User, I should receive a risk level (low, moderate, or high) based on my assessment responses.
As a User, I should see tailored recommendations based on my risk level.
As a User, I should see a clear warning message if my risk level is high.
As a User, I should be able to log my daily mood using a 1–5 emoji scale and add an optional note.
As a User, I should be able to view a dashboard with a trend graph of my mood history.
As a User, I should be able to view past assessment results on the dashboard.
As a User, I should see a medical disclaimer on every page of the website.

Page 3 of 6

4. User Personas

Anonymous User:
- Can complete the mental health assessment.
- Can log daily moods and view mood trends.
- Cannot save data across devices (data stored locally in the browser).

5. Visuals Colors and Theme

The amber-system will follow a custom "Midnight Zen" dark theme to create a calming and modern aesthetic. The design tokens are as follows:

primary: #1E293B (Deep Navy)
primary_light: #334155 (Slate Blue)
secondary: #F59E0B (Amber Gold)
accent: #10B981 (Emerald Green)
highlight: #F97316 (Warm Orange)
bg: #0F172A (Dark Blue)
surface: rgba(31, 41, 55, 0.9) (Semi-transparent Charcoal)
text: #F8FAFC (Light Gray)
text_muted: #94A3B8 (Muted Slate)
border: rgba(148, 163, 184, 0.2) (Soft Slate Border)

6. Signature Design Concept

The amber-system homepage will feature an interactive "Mindful Night Sky" experience.

Page 4 of 6

Concept Details:

Background: A dynamic, starry night sky with constellations that subtly shift and twinkle.
Interactive Elements:
- Users can hover over constellations to reveal motivational quotes or mental health tips.
- Clicking on a star will smoothly scroll the page to the corresponding section (e.g., "Start Assessment," "Mood Tracker," "About Us").
Animation:
- Smooth parallax scrolling with stars moving at different speeds using gsap and ScrollTrigger.
- Subtle hover effects on stars using framer-motion.
Transitions:
- Sections fade in and out as users scroll through the page.
- Buttons and links have a springy, tactile feel when clicked.

This concept will create a serene and engaging first impression, encouraging users to explore the website while maintaining a calming atmosphere.

7. Interaction Model & Motion Direction

Landing Page: Parallax interaction model with layered depth and scroll-triggered animations.
Internal Pages: Animated interaction model with hover transitions and smooth reveals for key elements.
Dashboard: Static interaction model for clarity and ease of use.

Each section of the landing page will have a unique interactive mechanic, such as hoverable constellations, scroll-triggered fades, and animated transitions.

Page 5 of 6

8. Non-Functional Requirements

The website must load within 3 seconds on a standard broadband connection.
The system must handle up to 1,000 concurrent users.
All inputs must be validated to prevent invalid or missing data.
The website must not collect or store personally identifiable information (PII).
The system must prioritize user safety by not underestimating risk levels.

9. Tech Stack

Frontend:
- React (Vite)
- TailwindCSS
- shadcn (UI components)
Backend:
- FastAPI
- Motor (MongoDB driver)
Database:
- MongoDB
Machine Learning:
- Logistic Regression (sklearn)
Testing:
- Cypress (frontend end-to-end tests)
- Pytest (backend unit tests)

Page 6 of 6

10. Assumptions and Constraints

The system assumes users have access to a modern web browser with JavaScript enabled.
Data is stored locally in the browser for anonymous sessions; no user accounts are supported.
The machine learning model is pre-trained using a placeholder dataset and may require updates for improved accuracy.
The website is designed for desktop and mobile browsers.

11. Glossary

PII: Personally Identifiable Information.
Rule-based Scoring: A method of calculating risk levels based on predefined rules.
Logistic Regression: A machine learning algorithm used for binary or multi-class classification.
Parallax: A design technique where background elements move slower than foreground elements, creating a sense of depth.
Recharts: A charting library for React used to create data visualizations.

This document outlines the complete system requirements for the amber-system website. Let me know if you need further refinements or additional details!