DevQuest.IO - Recommendation Service

Overview

The Recommendation Service is a component of the DevQuest.IO platform, responsible for generating personalized coding question suggestions for users. It leverages AI-driven insights, user progress data, and advanced data retrieval techniques to provide tailored recommendations.

---

Key Features

• Personalized Recommendations:
  • Suggests coding questions based on user progress and goals.
  • Prioritizes problems tagged similarly to previously solved questions.
• Data Integration:
  • Fetches and normalizes user progress data from external APIs.
  • Stores data in MongoDB for efficient querying and analysis.
• AI-Powered Insights:
  • Utilizes Retrieval-Augmented Generation (RAG) architecture with OpenAI APIs.
  • Combines embeddings from ChromaDB for context-aware suggestions.

---

Tech Stack

• Programming Language: Python
• Framework: FastAPI
• Database: MongoDB
• Embedding Storage: ChromaDB

---

Installation

Prerequisites

• Python 3.8+
• MongoDB instance (local or cloud)
• ChromaDB setup

Steps

1. Clone the repository:

   git clone <repository-url>
   cd recommendation-service

2. Create a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Configure environment variables:
   • Set up the following variables in a .env file:

     MONGO_URI=<your_mongodb_connection_string>
     OPENAI_API_KEY=<your_openai_api_key>
     CHROMADB_URI=<your_chromadb_connection_string>

---

Usage

1. Start the FastAPI server:

   uvicorn main:app --reload

2. API Endpoints:
   • GET /recommendations: Fetch personalized recommendations for a user.
     • Query Parameters:
       • user_id (string): Unique identifier for the user.
   • POST /sync-progress: Sync progress data from external APIs.
     • Body:

       {
         "user_id": "<user_id>",
         "platform_data": {
           "leetcode": { ... },
           "geeksforgeeks": { ... }
         }
       }

---

Architecture

1. Fetching Data:
   • Connects to external APIs to gather user progress data.
   • Normalizes and stores data in MongoDB.
2. Recommendation Generation:
   • Retrieves user embeddings from ChromaDB.
   • Queries OpenAI's API to enhance recommendations.
   • Suggests questions based on user goals and historical performance.