Skip to content

Latest commit

 

History

History

promptflow-autogen

README.md

What is Promptflow

Promptflow is a comprehensive suite of tools that simplifies the development, testing, evaluation, and deployment of LLM based AI applications. It also supports integration with Azure AI for cloud-based operations and is designed to streamline end-to-end development.

Refer to Promptflow docs for more information.

Quick links:

  • Why use Promptflow - Link
  • Quick start guide - Link

Getting Started

  • Install required python packages

    cd samples/apps/promptflow-autogen
pip install -r requirements.txt

  • This example assumes a working Redis cache service to be available. You can get started locally using this guide or use your favorite managed service

Chat flow

Chat flow is designed for conversational application development, building upon the capabilities of standard flow and providing enhanced support for chat inputs/outputs and chat history management. With chat flow, you can easily create a chatbot that handles chat input and output.

Create connection for LLM tool to use

You can follow these steps to create a connection required by a LLM tool.

Currently, there are two connection types supported by LLM tool: "AzureOpenAI" and "OpenAI". If you want to use "AzureOpenAI" connection type, you need to create an Azure OpenAI service first. Please refer to Azure OpenAI Service for more details. If you want to use "OpenAI" connection type, you need to create an OpenAI account first. Please refer to OpenAI for more details.

# Override keys with --set to avoid yaml file changes

# Create Azure open ai connection
pf connection create --file azure_openai.yaml --set api_key=<your_api_key> api_base=<your_api_base> --name open_ai_connection

# Create the custom connection for Redis Cache
pf connection create -f custom_conn.yaml --set secrets.redis_url=<your-redis-connection-url> --name redis_connection_url
# Sample redis connection string rediss://:PASSWORD@redis_host_name.redis.cache.windows.net:6380/0

Note in flow.dag.yaml we are using connection named aoai_connection for Azure Open AI and redis_connection_url for redis.

# show registered connection
pf connection show --name open_ai_connection

Please refer to connections document and example for more details.

Develop a chat flow

The most important elements that differentiate a chat flow from a standard flow are Chat Input, Chat History, and Chat Output.

  • Chat Input: Chat input refers to the messages or queries submitted by users to the chatbot. Effectively handling chat input is crucial for a successful conversation, as it involves understanding user intentions, extracting relevant information, and triggering appropriate responses.

  • Chat History: Chat history is the record of all interactions between the user and the chatbot, including both user inputs and AI-generated outputs. Maintaining chat history is essential for keeping track of the conversation context and ensuring the AI can generate contextually relevant responses. Chat History is a special type of chat flow input, that stores chat messages in a structured format.

    • NOTE: Currently the sample flows do not send chat history messages to agent workflow.

  • Chat Output: Chat output refers to the AI-generated messages that are sent to the user in response to their inputs. Generating contextually appropriate and engaging chat outputs is vital for a positive user experience.

A chat flow can have multiple inputs, but Chat History and Chat Input are required inputs in chat flow.

Interact with chat flow

Promptflow supports interacting via vscode or via Promptflow CLI provides a way to start an interactive chat session for chat flow. Customer can use below command to start an interactive chat session:

pf flow test --flow <flow_folder> --interactive

Autogen State Flow

Autogen State Flow contains stateflow example shared at StateFlow with Promptflow. All the interim messages are sent to Redis channel. You can use these to stream to frontend or take further actions. Output of Prompflow is summary message from group chat.

Agent Nested Chat

Autogen Nested Chat contains Scenario 1 of nested chat example shared at Nested Chats with Promptflow. All the interim messages are sent to Redis channel. You can use these to stream to frontend or take further actions. Output of Prompflow is summary message from group chat.

Redis for Data cache and Interim Messages

Autogen supports Redis for data caching and since redis supports a pub-subs model as well, this Promptflow example is configured for all agent callbacks to send messages to a Redis channel. This is optional feature but is essential for long running workflows and provides access to interim messages for your frontend. NOTE: Currently Promtpflow only support SSE for streaming data and does not support websockets. NOTE: In multi user chat bot environment please make necessary changes to send messages to corresponding channel.