AI Chatbot Web App with Spring Boot, Streaming Chat and Multi-LLM Routing

PHP, Java, JavaScript, PostgreSQL, Vue.js, Spring Boot, AI Chatbot Development, AI Chatbot, AI Model Integration, AI Automation · I want a production-ready AI chatbot that works as a personal assistant, streaming each token to the browser in real time. The core must be a clean Spring Boot REST backend that exposes an SSE endpoint, backed by PostgreSQL so every conversation—including its topic-based context—survives restarts.

The router should talk to multiple OpenAI-compatible LLMs (OpenAI, DeepSeek, Qwen, or any other endpoint I configure) and automatically fall back if one of them fails.

Image and file uploads are required. When a user drops a picture or PDF, the system should first attempt image captioning through a vision model and fall back to OCR when necessary so the chat can discuss the content.

On the frontend, a lightweight Vue 3 interface with chat bubbles and scrollable history is enough, as long as it reflects the streaming replies smoothly and lets the user revisit past sessions.

Deliverables • Spring Boot project with layered, readable code • Vue 3 client bundled in the same repo or a separate folder, your choice • Dockerfile(s) and docker-compose so I can spin everything up with one command • PostgreSQL schema and migration scripts • Setup README that lets me run, test, and ext