CollabWorks

CollabWorks is a Flask-based web application designed to connect clients with freelancers through role-based discovery and direct communication. The application provides authentication for two distinct user types, internal messaging, role prediction using a machine learning model, and basic profile discovery.

The system is implemented as a monolithic Flask application with modular routing and server-rendered templates.

Application Overview

The application supports two primary user roles:

• Clients
• Freelancers

Users authenticate via Flask-Login and interact with the platform through role-specific dashboards and shared messaging infrastructure.

Core Components

Authentication and User Management

• Uses Flask-Login for session management
• Supports multiple user models (Client, Freelancer)
• Dynamic user loading based on user ID and model type
• Role-based access control enforced at route level

Messaging System

• Conversation-based internal chat system
• Conversations are grouped using a conv_id
• Messages are persisted using SQLAlchemy
• Supports:
  • Client → Freelancer conversations
  • Freelancer → Client conversations
• Message metadata includes sender, receiver, timestamp, and ownership flag

Message Model

• conv_id identifies a conversation thread
• Messages are retrieved and rendered per user context
• Server-generated responses supported for testing/demo purposes

Chat Interfaces

• Separate chat views for clients and freelancers
• Conversation lists are dynamically built from stored messages
• Active conversation context is rendered server-side
• Messages can also be fetched asynchronously via JSON endpoints

Role Prediction (Machine Learning)

• Uses pre-trained models loaded via joblib
• Models include:
  • Multi-label classifier
  • Label binarizer
  • Per-role probability thresholds
• Accepts free-text need statements
• Predicts up to N relevant freelancer roles
• Filters low-confidence and generic predictions
• Persists prediction history to a JSON file for later retrieval

Freelancer Discovery

• Exposes a JSON endpoint for retrieving freelancer profiles
• Profiles include:
  • Name
  • Username
  • Tagline
  • Location
  • Hourly rate (derived)
  • Rating (derived)
• Profile images are assigned dynamically based on gender
• Intended for frontend-driven listing and search

Client Status Validation

• Provides a lightweight endpoint to verify client authentication state
• Used for frontend gating and conditional rendering

Routing Overview

The application exposes routes for:

• Authentication (via registered blueprints)
• Messaging and chat navigation
• Role prediction and retrieval
• Freelancer discovery
• Static informational pages
• Dashboard rendering

Blueprints are used to separate:

• Client-specific routes
• Freelancer-specific routes

Technology Stack

Backend

• Python 3.12
• Flask
• Flask-Login
• Flask-SQLAlchemy
• Flask-Bcrypt
• SQLite (primary datastore)
• Joblib (ML model loading)

Frontend

• Server-rendered HTML templates (Jinja2)
• Vanilla JavaScript
• Static CSS assets

Data Storage

• SQLite database used via SQLAlchemy ORM
• Direct SQLite queries used for specific dashboard views
• JSON file used for persisting role prediction history

Execution Model

• Runs as a single Flask application
• Synchronous request handling
• ML inference performed inline during request processing

Notes

• Client and Freelancer are treated as separate user models
• Messaging logic assumes a shared conversation namespace
• Role prediction is best-effort and threshold-based
• The application is designed for incremental extension rather than horizontal scaling