https://github.com/roelium/hotel-ai-assistant-backend
The backend part of a fullstack simple hotel booking management system with AI assistant that can interact with booking data. For educational/experimental purpose to show case the usage of Langchain4J in interacting with java functions through its LLM tooling capabilities.
https://github.com/roelium/hotel-ai-assistant-backend
aws-s3 java jwt-token langchain4j maven mysql mysql-database openai sping-boot
Last synced: 2 months ago
JSON representation
The backend part of a fullstack simple hotel booking management system with AI assistant that can interact with booking data. For educational/experimental purpose to show case the usage of Langchain4J in interacting with java functions through its LLM tooling capabilities.
- Host: GitHub
- URL: https://github.com/roelium/hotel-ai-assistant-backend
- Owner: roelium
- Created: 2025-03-15T15:37:37.000Z (3 months ago)
- Default Branch: master
- Last Pushed: 2025-04-08T13:53:44.000Z (2 months ago)
- Last Synced: 2025-04-11T21:17:59.501Z (2 months ago)
- Topics: aws-s3, java, jwt-token, langchain4j, maven, mysql, mysql-database, openai, sping-boot
- Language: Java
- Homepage:
- Size: 43.8 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Hotel - AI Assistant - backend
The backend part of a fullstack simple hotel booking management system with AI assistant that can interact with booking data. For educational/experimental purpose to show case the usage of Langchain4J in interacting with java functions through its AI LLM tooling capabilities.
Medium article for more context: [https://medium.com/@syahrulanw/integrating-an-ai-assistant-with-rag-using-langchain4j-in-a-spring-boot-react-js-17fdc76676b4](https://medium.com/@syahrulanw/integrating-an-ai-assistant-with-rag-using-langchain4j-in-a-spring-boot-react-js-17fdc76676b4)
## Tech stack
Frontend: React, Javascript, Typescript
Backend: Java, Spring Boot, MySQL, Maven, AWS S3 bucket, JWT token, Langchain4j, Open AI API
## Feature
common: login, register, find available rooms for certain dates
guest: book a room, view bookings, cancel a booking, find a booking, view/edit profile
admin: add/edit room, manage bookings
The AI asistant currently can find and cancel a booking for a guest.
## Steps to Run the Backend
1. Clone the project
2. Configure the application.properties file with:
- A local or remote MySQL connection. Create a new database instance (without tables).
- Create a new or use an existing AWS S3's access key, secret key and a bucket.
- Createa new or use an existing Open AI Key on langchain4j.open-ai.chat-model.api-key and langchain4j.open-ai.embedding-model.api-key. You can use "demo" as key for free provided by LangChain4j but i haven't try.
3. Run the RoelHotelApplication with Maven (`./mvnw spring-boot:run`).
4. Use API tool like Postman to try the API endpoints. The AI chat functionality requires an authorized bearer token (users need to be registered and loggedin first). Request examples:
````
```
endpoint: POST localhost:4040/auth/register
body:
{
"email":"[an email address]",
"password":"[password]",
"phoneNumber":"12345678",
"name":"Guest 1",
"role":"USER"
}
endpoint: POST localhost:4040/auth/login
body:
{
"email":"[an email address]",
"password":"[password]"
}
endpoint: GET localhost:4040/chat?chatid=1&message="could you please get booking details of 9AZ951P9DQ".
Auth type: bearer token from the login step
```
````
## Important Parts for using the LangChain4J
1. Declare dependencies (file: pom.xml)
```xml
dev.langchain4j
langchain4j-spring-boot-starter
${langchain4j.version}
dev.langchain4j
langchain4j-open-ai-spring-boot-starter
${langchain4j.version}
dev.langchain4j
langchain4j-embeddings-all-minilm-l6-v2
${langchain4j.version}
dev.langchain4j
langchain4j-reactor
${langchain4j.version}
```
2. Declare an AI service for the role it is playing (file: LangChain4jAssistant.java)
```java
@AiService
public interface LangChain4jAssistant {
@SystemMessage("""
You are a customer chat support agent of a Hotel named "Roel Hotel".
Respond in a friendly, helpful, and joyful manner.
You are interacting with customers through an online chat system.
Before providing information about a booking or cancelling a booking.
Check the message history for this information before asking the user.
Before changing a booking, you MUST ensure it is permitted by the terms.
If there is a charge for the change, you MUST ask the user to consent before proceeding.
Use the provided functions to fetch booking details and cancel bookings.
Today is {{current_date}}.
""")
String chat(@MemoryId String chatId, @UserMessage String userMessage);
}
```
3. Define LangChain4j configurations like chatMemoryProvider (important so that the asistant can remember previous messages), embeddingStore, contentRetriever (file: LangChain4jConfig.java)
4. Define LangChain4j tools. This is where the LangChain4J can find the available tools/functions to be called based on chat interaction. (file: LangChain4jTools.java)
5. Publish the API endpoint for sending the chat message (file: ChatController.java)
6. Optional: to give additional context to the AI service from a document or file, e.g. term of service, we can use document ingestor implementation (files: DocumentationIngestor.java & terms-of-service.txt).
## References and Inspirations
1. https://docs.langchain4j.dev/tutorials/spring-boot-integration/
2. https://github.com/marcushellberg
3. https://github.com/achalise
4. https://github.com/phegondev
## Screenshots
