This version is functionally the same as my AI-102 console practice project of January this year, but with a few important changes:
- Instead of two local LLM agents running via Ollama, it now uses two Cloudflare Qwen2.5 LLM agents.
- Application is containerized and deployed to Azure Container Apps.
- The page is served by a Cloudflare Worker.
- Output is streamed in real time to the web page with SignalR.
The code is managed in GitHub and deployed automatically through a GitHub Actions workflow.
Planned next steps:
Version 2: let the user enter a word to be guessed. Version 3: let the user play against an LLM agent.
Try it here: llm-hangman-demo.koorevaar.com