How henryxiloj/demo-ollama-deepseek-r1 works
Project Type and Purpose
• This is a full-stack application focused on demonstrating local AI inference.
• The backend, powered by Spring Boot 3.4.2/Java 21, exposes a chat API endpoint that integrates with Ollama’s local LLM (DeepSeek-R1).
• The frontend is an Angular application that provides a chat interface to interact with the AI service.
• Its use case emphasizes local processing for privacy and low latency without reliance on cloud APIs.
File Structure Analysis
• The top-level directories clearly separate the Angular frontend (in the “angular-ui” folder) and the backend source (under “src/main/java/com/henry/ollama”).
• The “angular-ui” folder contains standard Angular configuration (angular.json, package.json) and source files under “angular-ui/src/app” that include components like “chat.component”.
• The backend structure includes the Application entry point, configuration files (OllamaProperties), controllers (ChatController, WebController), service logic (OllamaService), and domain records (OllamaRequest, OllamaResponse).
• Maven build files (pom.xml, mvnw, etc.) indicate that the project builds the full-stack solution, including frontend asset installation and compilation.
Practical Steps
• Within the backend, break out the Controller, Service, and Configuration components.