- LangChain, LangServe, FastAPI, and OpenAI API used for backend.
- React.js, HTML, and CSS used for frontend.
- Elasticsearch database as external vectorstore.
- Agent acts as Romeo from Romeo and Juliet.
- Deployed backend using Railway.app and frontend using GitHub Pages.
LLM Agent Tools:
- Retrieval tool: for LLM Agent to access all of Romeo's lines from the original script. Agent uses this tool to answer questions about Romeo.
- Web Search tool: Tavily Search API used to search information our LLM Agent does not know (including current events.)
- Weather tool: accessed through weather API, to find current weather conditions in a city.
- Math tool: used to answer math questions.
- YouTube search tool: used to find YouTube video URLs.
Preview: Watch demo
Backend | Frontend |
---|---|
Retrieval tool: |
|
Weather tool: |
|
Web Search tool: |
|
Math tool: |
|
YouTube tool: |