feat: Persistent chat state (#181)

* add memory saver and update api contract

* update new api contract in frontend and chat widget

* add mongo memory saver

* update read me

Author: alfredfrancis

README.md

@@ -3,15 +3,31 @@
 [![Join the chat at https://gitter.im/ai-chatbot-framework/Lobby](https://badges.gitter.im/ai-chatbot-framework/Lobby.svg)](https://gitter.im/ai-chatbot-framework/Lobby?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge) [![Build Status](https://github.com/alfredfrancis/ai-chatbot-framework/actions/workflows/evaluate-backend.yml/badge.svg)](https://github.com/alfredfrancis/ai-chatbot-framework/actions/workflows/evaluate-backend.yml) [![Build Status](https://github.com/alfredfrancis/ai-chatbot-framework/actions/workflows/evaluate-frontend.yml/badge.svg)](https://github.com/alfredfrancis/ai-chatbot-framework/actions/workflows/evaluate-frontend.yml)
 
 
+AI Chatbot Framework is an open-source, self-hosted, DIY Chatbot building platform built in Python. With this tool, it’s easy to create Natural Language conversational scenarios with no coding efforts whatsoever. 
+The smooth UI makes it effortless to create and train conversations to the bot. AI Chatbot Framework can live on any channel of your choice (such as Messenger, Slack etc.) by integrating it’s API with that platform.
 
-### An AI Chatbot framework built in Python
+You don’t need to be an expert at artificial intelligence to create an awesome chatbot that has AI capabilities. With this project you can create an AI powered chatting machine in no time. 
 
-AI Chatbot Framework is an AI powered conversational dialog interface built in Python. With this tool, it’s easy to create Natural Language conversational scenarios with no coding efforts whatsoever. The smooth UI makes it effortless to create and train conversations to the bot. AI Chatbot Framework can live on any channel of your choice (such as Messenger, Slack etc.) by integrating it’s API with that platform.
-
-You don’t need to be an expert at artificial intelligence to create an awesome chatbot that has AI capabilities. With this boilerplate project you can create an AI powered chatting machine in no time. Since this is a hobby project, there could be numerous bugs, so contributions through pull requests are welcome!
 
 ![](docs/screenshots/admin_chat_screenshot.png)
 
+## Features
+- Fully Self Hosted
+- Multi-turn Conversations
+- Advanced Natural Language Understanding (NLU)
+  - Spacy Word Embeddings
+  - Intent Recognition (ML)
+  - Entity Extraction (ML)
+  - One shot NLU using Large Language Models (Coming Soon)
+- Persistent Memory & Context Management
+- API request fulfilment
+- Channel Integrations
+  - Web using rest api/chat snippet
+  - Facebook Messenger (coming soon)
+  - WhatsApp via Twilio (coming soon)
+  - Slack (coming soon)
+- Low Code Admin Dashboard
+
 ## Index
 
 * [Installation](#installation)

app/admin/test/__init__.py

@@ -0,0 +1,24 @@
+from fastapi import APIRouter, Depends
+from app.bot.dialogue_manager.models import UserMessage
+from app.dependencies import get_dialogue_manager
+from app.bot.dialogue_manager.dialogue_manager import DialogueManager
+
+router = APIRouter(prefix="/test", tags=["test"])
+
+
+@router.post("/chat")
+async def chat(
+    body: dict, dialogue_manager: DialogueManager = Depends(get_dialogue_manager)
+):
+    """
+    Endpoint to converse with the chatbot.
+    Delegates the request processing to DialogueManager.
+
+    :return: JSON response with the chatbot's reply and context.
+    """
+
+    user_message = UserMessage(
+        thread_id=body["thread_id"], text=body["text"], context=body["context"]
+    )
+    new_state = await dialogue_manager.process(user_message)
+    return new_state.to_dict()

app/admin/test/routes.py

@@ -1,13 +1,14 @@
-from fastapi import APIRouter, Request, Depends
-from app.bot.dialogue_manager.models import ChatModel
+from fastapi import APIRouter, Depends
+from app.bot.dialogue_manager.models import UserMessage
 from app.dependencies import get_dialogue_manager
+from app.bot.dialogue_manager.dialogue_manager import DialogueManager
 
 router = APIRouter(prefix="/v1", tags=["bots"])
 
 
 @router.post("/chat")
 async def chat(
-    request: Request, body: dict, dialogue_manager=Depends(get_dialogue_manager)
+    body: dict, dialogue_manager: DialogueManager = Depends(get_dialogue_manager)
 ):
     """
     Endpoint to converse with the chatbot.
@@ -16,7 +17,8 @@ async def chat(
     :return: JSON response with the chatbot's reply and context.
     """
 
-    # Access the dialogue manager from the fast api application state.
-    chat_request = ChatModel.from_json(body)
-    chat_response = await dialogue_manager.process(chat_request)
-    return chat_response.to_json()
+    user_message = UserMessage(
+        thread_id=body["thread_id"], text=body["text"], context=body["context"]
+    )
+    new_state = await dialogue_manager.process(user_message)
+    return new_state.bot_message