Ralph's Blog about this and that (RSS | Atom)

Web Search with LLM in Go

Published: 28. October 2024  •  llm, go

Combine web search with an LLM in Go and query local models running on Ollama.

Running an LLM in the browser with Transformers.js

Published: 27. October 2024  •  ionic, angular, llm

Run an LLM in a web application with Transformers.js.

Fine-tune a LLM

Published: 27. October 2024  •  python, llm

Fine-tune a small language model for SQL generation and convert it for use with Transformers.js.

Running LLM-Generated Go Code in a Docker Container

Published: 17. January 2025  •  Updated: 10. September 2025  •  go, llm

In this article, I will show you how to write a Go program that generates Go code with the help of a large language model (LLM) and run this generated code in a Docker container.

Tool/Function Calling with Spring AI

Published: 29. January 2025  •  java, spring, llm

A look at Spring AI's support for Tool/Function calling. How to write your own tools and integrate them into your Spring AI application.

Speech Recognition in the Browser with Transformers.js

Published: 19. January 2025  •  ionic, angular, llm

Running the Moonshine model, a speech recognition model, in the browser with Transformers.js. The demo application is a snake game controlled by voice commands.

RAG with Spring AI

Published: 23. July 2025  •  spring, llm

Implementing a traditional RAG (Retrieval Augmented Generation) system using Spring AI.

Model Context Protocol (MCP) with Go

Published: 1. December 2025  •  go, llm

Learn how to write MCP servers and clients in Go to build interoperable LLM tools using the Model Context Protocol (MCP).

Code Mode for Tool Calling in Go

Published: 11. March 2026  •  llm, go

Tool calling is a workflow for giving an LLM access to external information and capabilities. However, it can be inefficient when you have a large number of tools or complex workflows that require multiple dependent tool calls. This blog post explores code mode, a way to use tools that can save context tokens and reduce the number of request/response cycles by incorporating a tool search and letting the model write code that orchestrates tool calls. The blog post shows you how to implement code mode yourself in Go.

Code Mode for Tool Calling in Java

Published: 7. April 2026  •  llm, java

Code mode is a way to use tools that can save context tokens and reduce the number of request/response cycles by incorporating a tool search and letting the model write code that orchestrates tool calls. This blog post explores the code mode pattern and shows how to implement it in Java.

Eino: A Go Framework for Building LLM Applications

Published: 12. April 2026  •  go, llm

Eino is a Go-based framework for building LLM applications, from simple chat features to agent systems with tools, middleware, and orchestration. This post explores Eino's core abstractions and patterns through code examples.