Chatbot Builder Executor

Description

This is the execution service for the Chatbot Builder, responsible for executing the chatbot flow logic such as LLM generation calls, and other processing tasks.

Setup

1. Add .env file in the app directory with variables:

   • OPENAI_ENDPOINT
   • OPENAI_KEY

2. If you want to use DeepSeek Model instead of the default AzureOpenAi model, add the following variables too:

   • USE_DEEPSEEK and set it to true
   • DEEPSEEK_ENDPOINT and set it to https://api.deepseek.com
   • DEEPSEEK_KEY

3. Before running the service, you’ll need to generate the gRPC code from the .proto files to ensure the service can communicate with other services.

Generate gRPC Code Locally

To generate the necessary gRPC code for Python, follow these steps:

1. Clone or Pull the Latest .proto Files

   git clone https://github.com/Chatbot-Builder-Project/chatbot-builder-protos.git

   Or if you already have the repository:

   cd chatbot-builder-protos
   git pull
   cd ..

2. Create Directory Structure

   mkdir -p app/generated/v1

3. Generate Python gRPC Code

   python -m grpc_tools.protoc --proto_path=chatbot-builder-protos/protos --python_out=app/generated --grpc_python_out=app/generated chatbot-builder-protos/protos/v1/executor/*.proto

4. Fix Imports

   Get-ChildItem -Path "app/generated/v1/executor/" -Filter "*_pb2*.py" | ForEach-Object {
    (Get-Content $_.FullName) -replace "from v1\.executor", "from app.generated.v1.executor" | Set-Content $_.FullName
   }

Commands

Build & Run

docker-compose up --build