codefarm0/ai-powered-text-summariser

This project is a design document and reference build for an AI-powered text summarizer written with Spring Boot on the backend and React on the frontend. Rather than a quick demo of sending one prompt to an AI model, it walks through designing a full application the way a real engineering team would, covering how the pieces fit together before any code is written. The application lets people organize their AI work inside Projects. Instead of generating a summary in isolation, every document upload, summary, and conversation belongs to a specific project, similar to how modern note taking or AI tools group related work together. Within a project, a user can upload PDF, DOCX, or TXT files, generate different kinds of summaries such as short, detailed, executive, bullet point, or technical versions, and then keep chatting with the AI to ask follow up questions, get explanations, generate quizzes, or get translations. Every interaction is saved automatically so users can revisit past prompts, summaries, and conversations. To keep the focus on learning AI concepts rather than building a login system, the app starts with a lightweight guest user setup. On a person's first visit, the browser generates a unique identifier and stores it locally, and the backend uses that to automatically create a guest account tied to that browser. There is no sign in step, but a person's projects and history stick around on return visits from the same browser. The plan is to later swap this out for proper authentication without changing the rest of the application. Everything runs on local, open source language models through Ollama, including options like Mistral, Llama 3, and Qwen, so there is no need for a paid API subscription. The technology stack includes Java 21, Spring Boot, Spring AI, and MySQL on the backend, with React, Tailwind CSS, and Axios on the frontend. Spring AI handles the connection to Ollama, including prompt templates, streaming responses, and structured output, while Spring Boot manages the business logic, REST APIs, and saving data to MySQL. The full README is longer than what was shown.