🤖 TechTorque Unified AI Agent / RAG Service

The Agent_Bot service is the intelligence and interaction layer of the TechTorque platform. It combines two critical AI functionalities:

1. AI Agent (Tool Use): Uses advanced reasoning to perform real-time actions against microservices (e.g., checking appointment slots, viewing user vehicle status).
2. RAG (Knowledge Retrieval): Uses a Vector Database (Pinecone) and a local embedding model to answer non-real-time questions based on structured knowledge documents (e.g., "What is your warranty policy?").

Built with Python, FastAPI, and Gemini (via LangChain), this service implements the /api/v1/ai/chat endpoint.

---

🚀 Setup and Local Development

Prerequisites

• Python 3.10+
• External Microservices: The Authentication, Appointment_Service, Vehicle_Service, and Project_Service must be running.
• Cloud Services: A Gemini API Key and a Pinecone Account/API Key are required for RAG functionality.

1. Environment Setup

1. Navigate to the project directory:

   cd Agent_Bot

2. Activate the virtual environment:

   source venv/bin/activate  # macOS/Linux
   .\venv\Scripts\activate    # Windows

3. Install dependencies: (Ensure you run this command inside the active (venv) to resolve all LangChain dependencies)

   pip install -r requirements.txt 

2. Configuration (.env)

Create a .env file in the root of the Agent_Bot directory and populate it with your specific secrets and URLs.

# --- LLM & RAG Configuration ---
GOOGLE_API_KEY="YOUR_ACTUAL_GEMINI_API_KEY_HERE"
GEMINI_MODEL="gemini-2.5-flash"

PINECONE_API_KEY="YOUR_ACTUAL_PINECONE_API_KEY_HERE"
PINECONE_ENVIRONMENT="us-east-1-aws"
PINECONE_INDEX_NAME="techtorque-kb"

# RAG Configuration Defaults
RAG_CHUNK_SIZE=500
RAG_CHUNK_OVERLAP=50
MAX_CONTEXT_LENGTH=2000

# --- Microservice URLs ---
PORT=8091
BASE_SERVICE_URL="http://localhost:8080/api/v1" 

AUTHENTICATION_SERVICE_URL="${BASE_SERVICE_URL}/auth"
VEHICLE_SERVICE_URL="${BASE_SERVICE_URL}/vehicles"
PROJECT_SERVICE_URL="${BASE_SERVICE_URL}/jobs" 
TIME_LOGGING_SERVICE_URL="${BASE_SERVICE_URL}/logs"
APPOINTMENT_SERVICE_URL="${BASE_SERVICE_URL}/appointments"