About:

Dan Corin is a software engineer interested in language models, who enjoys experimenting and building small projects.

Website:

Specializations:

Interests:

Language models

Incoming Links:

Subscribe to RSS:
The blog post discusses the use of DSPy, a library for running inference with language models, through the example of counting letters in the word 'strawberry'. It explains the concept of a 'Signature' in DSPy, which structures in...
The blog post explores the differences in streaming performance among various language models, specifically focusing on OpenAI's models, Gemini from Google, and Anthropic's Claude. The author shares insights from personal experime...
The blog post explores the differences in streaming performance among various language models (LLMs) including OpenAI's GPT-3.5 and GPT-4, Gemini's models, and Anthropic's Claude. The author shares insights from experiments conduc...
In 2025, the author explores the transformative impact of coding agents on software development, emphasizing collaboration, continuous learning, and personal project advancements.
The blog post discusses the author's experience with the sqlite-vector project, focusing on storing and querying embedding vectors in a SQLite database. The author details the process of setting up the environment, installing nece...
The blog post discusses the author's experimentation with DSPy, a lightweight framework for working with large language models (LLMs). It focuses on understanding token usage, reasoning, and prompt configurations when extracting d...
The blog post discusses the author's experience with Model Context Protocol (MCP) servers, particularly focusing on building a simple MCP server using the fastmcp library. The author details the process of setting up the server, r...
The blog post discusses the author's experience setting up a Model Context Protocol (MCP) server using the fastmcp library to expose a tool for generating events based on date inputs. The author shares insights on the challenges f...
The blog post discusses the author's experience setting up a Model Context Protocol (MCP) server using the fastmcp library to expose a tool for generating events based on date inputs. The author shares insights on the challenges f...
The post explores the differences between writing and speaking, emphasizing that writing is a nonlinear process that allows for deeper thinking and refinement of ideas. It discusses the mental bandwidth required for both activitie...
The blog post discusses the author's experiences with the OpenAI Chat Completions API, highlighting its utility for testing various models and providers. It details a specific issue encountered while using a third-party provider t...
The blog post discusses the author's experiences with the OpenAI Chat Completions API, highlighting its utility for testing various models and providers. It details the author's setup of a Python app using the OpenAI client, the c...
The author discusses their experience migrating their blog, Thought Eddies, to the Astro framework. They detail the challenges faced during the migration, including slow build times and performance issues related to the implementa...
The author discusses their experience migrating their blog, Thought Eddies, to the Astro framework. They detail the challenges faced during the migration, including slow build times and performance issues related to generating Ope...
The blog post discusses the challenges non-engineers face when using coding agents like Cursor, particularly in managing side effects and utilizing checkpoints effectively. It contrasts the experience of software engineers, who ar...
The post explores the changing landscape of writing in software development, highlighting the collaboration with coding agents and the importance of storytelling in technical communication.
Images alone are insufficient for coding agents to produce pixel-perfect designs; detailed language descriptions enhance clarity and effectiveness in coding tasks.
The blog post discusses the challenges non-engineers face when using coding agents like Cursor, particularly in managing side effects and understanding the context of code changes. It highlights the importance of checkpoints for r...
The post discusses the use of the .env file and the python-dotenv package for managing environment variables in Python. It highlights the common confusion around installing the correct package and introduces the 'uv' tool as a sol...
The post discusses the use of the .env file and the python-dotenv package for managing environment variables in Python. It highlights the common confusion around installing the correct package and introduces the 'uv' tool as a sol...
The author discusses their experience using code agents to assist in coding tasks, particularly in creating a music playlist application that integrates with Spotify. They detail the process of using a language model to generate f...
The blog post discusses the author's experience using code agents to assist in coding tasks, particularly in creating a music playlist application that integrates with Spotify. The author outlines the process of using a language m...
The blog post discusses the functionality of custom slash commands in Claude Code, allowing users to define frequently-used prompts as Markdown files. It explains how commands can be organized by scope (project-specific or persona...
The author shares their experience training to run a mile in under 6 minutes using various language models to create a training plan. They reflect on the importance of personal knowledge in evaluating AI-generated fitness advice a...