System Requirements Document (SRD)

Project Name: ocean-ai

1. Introduction

The ocean-ai project is an AI-powered personal wardrobe assistant designed to help users maximize the potential of their existing clothing. By intelligently combining wardrobe items into stylish outfits, ocean-ai aims to solve the common "I have nothing to wear" problem. This system is tailored for individuals who want to dress with more intention, confidence, and personal style, without the need for additional shopping.

This document outlines the system requirements for ocean-ai, incorporating the newly requested feature of style profile functionality that allows users to select aesthetic preferences during onboarding and per outfit request.

2. System Overview

ocean-ai is a cloud-based application that leverages AI to analyze user-uploaded clothing photos, categorize them, and suggest outfit combinations based on user-defined style preferences. The system is designed to provide a seamless and intuitive experience, enabling users to build a digital wardrobe, define their aesthetic, and receive personalized outfit recommendations.

Key features include:

• Digital wardrobe creation through photo uploads.
• AI-driven categorization and tagging of clothing items.
• Style profile functionality for aesthetic customization.
• Outfit suggestions tailored to occasions and style moods.
• Pairing recommendations for individual wardrobe items.

The system will prioritize a clean, modern design with a focus on usability and visual appeal, ensuring that users in Azerbaijan (AZ) and beyond can easily navigate and benefit from the platform.

3. Functional Requirements

• As a User, I should be able to create an account and log in securely.
• As a User, I should be able to upload photos of my clothing items to build a digital wardrobe.
• As a User, I should be able to view my wardrobe items organized by categories (e.g., tops, bottoms, shoes).
• As a User, I should be able to select my preferred aesthetic styles (e.g., Old Money, Streetwear, Minimalist) during onboarding.
• As a User, I should be able to request a full outfit suggestion by selecting an occasion and/or style mood.
• As a User, I should be able to override my default style preferences for a specific outfit request.
• As a User, I should be able to tap on any wardrobe item to see AI-recommended pairings filtered by my style preferences.
• As a User, I should be able to update my style preferences at any time through the settings page.
• As a User, I should be able to confirm or adjust AI-suggested categories and tags for uploaded clothing items.

4. User Personas

4.1 Everyday User

• Description: Individuals who want to simplify their daily outfit choices and dress with more confidence.
• Goals: Save time, maximize wardrobe usage, and align outfits with personal style preferences.
• Pain Points: Struggles with outfit combinations, feels overwhelmed by wardrobe options.

4.2 Fashion Enthusiast

• Description: Style-conscious users who enjoy experimenting with different aesthetics and trends.
• Goals: Explore new styles, create unique combinations, and express individuality through fashion.
• Pain Points: Difficulty visualizing new combinations, wants to avoid repetitive outfits.

5. Visuals Colors and Theme

Color Palette

The ocean-ai platform will use a unique, modern color palette inspired by the ocean and fashion aesthetics:

• Background: #F4F9FC (Soft Ocean Mist)
• Surface: #FFFFFF (Clean White)
• Text: #2C3E50 (Deep Navy)
• Accent: #1ABC9C (Seafoam Green)
• Muted Tones: #BDC3C7 (Pale Gray)

This palette ensures a clean, minimalistic design that allows clothing photos to remain the focal point.

6. Signature Design Concept

The ocean-ai homepage will feature an interactive digital wardrobe closet.

• Visual Concept: The homepage will resemble a sleek, modern closet with sliding glass doors. As users scroll, the doors slide open to reveal their wardrobe items, displayed as high-quality, floating images.
• Animations: Each clothing item will subtly "hover" with a soft shadow effect, and tapping on an item will trigger a smooth zoom-in animation, transitioning to the Item Detail screen.
• Transitions: The background will shift between soft gradients of ocean-inspired colors (e.g., seafoam green to pale blue) as users interact with the interface.
• Micro-Interactions: When users select a style mood or occasion, the closet dynamically rearranges to highlight items that fit the selected criteria, creating a visually engaging experience.
• Mood: The design will evoke a sense of calm, organization, and luxury, making users feel like they have a personal stylist at their fingertips.

7. Non-Functional Requirements

• The system must support high-resolution image uploads and processing.
• The AI model should provide accurate categorization and style recommendations with a response time of under 2 seconds.
• The platform must ensure data security and privacy, particularly for user-uploaded photos.
• The application should be optimized for both web and mobile platforms, with responsive design.
• The system must support at least 10,000 concurrent users without performance degradation.

8. Tech Stack

Frontend

• React for Web
• React Native for Mobile App

Backend

• Python
• FastAPI

Database

• MySQL (with Alembic for migrations)

AI Models

• GPT 5.4 for user-friendly responses
• Claude 4.6 Opas for coding and academic tasks
• Google Nano Banana for image generation

AI Tools

• Litellm for LLM Routing
• Langchain

Orchestration

• Docker and docker-compose for local development
• Kubernetes for server-side orchestration

9. Assumptions and Constraints

• Users will upload clear, well-lit photos of their clothing items for accurate AI analysis.
• The system will initially support English as the primary language, with potential for localization based on demand.
• The platform will operate in the AZ timezone by default.
• Internet connectivity is required for all core functionalities.

10. Glossary

• AI: Artificial Intelligence, used for analyzing clothing photos and generating outfit suggestions.
• Style Profile: A user-defined set of aesthetic preferences (e.g., Old Money, Streetwear).
• Outfit Suggester: A feature that provides complete outfit recommendations based on user preferences.
• Item Pairing: A feature that suggests complementary wardrobe items for a selected piece.
• Digital Wardrobe: A virtual representation of a user's clothing collection.