Developing Chat Agents: Getting Started

What is a Chat Agent?

It's a function which calls Large Language Models (LLMs) to respond to the student's messages given contxtual data:

• question data
• user data such as past responses to the problem Chatbot Agents capture and automate the process of assisting students during their learning process when outside of classroom.

Getting Setup for Development

1. Get the code on your local machine (Using github desktop or the git cli)

   • For new functions: clone the main repo for lambda-chat and create a new branch. Then go under scr/agents and copy the base_agent folder.

   • For existing functions: please make your changes on a new separate branch

2. If you are creating a new chatbot agent, you'll need to set it's name as the folder name in scr/agents and its corresponding files.

3. You are now ready to start making changes and implementing features by editing each of the three main function-logic files:

   1. scr/agents/{base_agent}/{base}_agent.py: This file contains the main LLM pipeline using LangGraph and LangChain.

   2. the agent expects the following inputs when it being called:

   Body with necessary Params:

   {
       "message":"hi",
       "params":{
               "conversation_id":"12345Test",
               "conversation_history": [{"type":"user","content":"hi"}]
       }
   }
   

   Body with optional Params:

   {
       "message":"hi",
       "params":{
               "conversation_id":"12345Test",
               "conversation_history":[{"type":"user","content":"hi"}],
               "summary":" ",
               "conversational_style":" ",
               "question_response_details": "",
               "include_test_data": true,
               "agent_type": {agent_name}
       }
   }
   

4. scr/agents/{base_agent}/{base}_prompts.py: This is where you can write the system prompts that describe how your AI Assistant should behave and respond to the user.

5. Make sure to add your agent invoke() function to the module.py file.

6. Please add a README.md file to describe the use and behaviour of your agent.

7. Changes can be tested locally by running the pipeline tests using:

   pytest src/module_test.py
   

   Running and Testing Agents Locally

8. Merge commits into any branch (except main) will trigger the dev.yml workflow, which will build the docker image, push it to a shared dev ECR repository to make the function available from the dev and localhost client app.

9. In order to make your new chatbot available on the LambdaFeedback platform, you will have to get in contact with the ADMINS on the platform.