• POST Request: The client sends a POST request to /query with a JSON payload that includes:

  • session_id
  • text

  Example Payload:

  {"session_id": "user123", "text": "My laptop won’t turn on"}

• Validation: The Query model in models.py validates the input.

• Session Handling: persistence.py either loads the current session state or initializes a new one.
• History Update: The submitted query is appended to the session’s history.

The workflow.py module executes the LangGraph graph through the following steps:

• Query Ingestion: Adds the query to the state.
• NLP Processing: Invokes groq_client.py to extract structured information (e.g., issue, priority).
• Decision Making: Determines whether to auto-respond or escalate the query based on its priority/urgency.
• Response Generation: Constructs a response based on the decision made.

• State Saving: The updated state is saved.
• Client Response: The final response is returned to the client as a Response object.

Run the following command in your terminal:

pip install fastapi uvicorn groq langgraph pydantic

• Open config.py
• Replace "your_groq_api_key" with your actual Groq API key.

Start the app by executing:

python app.py

Send a test POST request using curl:

curl -X POST "http://localhost:8000/query" \
     -H "Content-Type: application/json" \
     -d '{"session_id": "user123", "text": "My laptop won’t turn on after I updated it yesterday."}'

The expected JSON response should appear as follows:

{
  "response": "We have identified your issue as 'won’t turn on'. Please try this solution: [Solution Placeholder]"
}