---
title: "Dan Girellini (@dangit) | Hashnode"
description: "Dan Girellini (@dangit) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@dangit"
---

@dangit

# Dan Girellini

@dangit · Joined August 2024

### [Your evals have a Rotten Tomatoes problem](https://engineering.fractional.ai/your-evals-have-a-rotten-tomatoes-problem?utm_source=hashnode&utm_medium=feed)

You push a change to a prompt and your eval score drops from 0.91 to 0.84. Something got worse, but the score doesn’t tell you what. So you start re-running the pipeline, tweaking your inputs, pouring

8 min read · [engineering.fractional.ai](https://engineering.fractional.ai/your-evals-have-a-rotten-tomatoes-problem?utm_source=hashnode&utm_medium=feed)

[Feb 12](https://hashnode.com/posts/your-evals-have-a-rotten-tomatoes-problem/698e4661cd9d95e59895acd5)

### [Taming LLM Responses: Dynamic Pydantic Models for Flexible Structured Output](https://engineering.fractional.ai/taming-llm-responses-dynamic-pydantic-models-for-flexible-structured-output?utm_source=hashnode&utm_medium=feed)

As developers working with Large Language Models (LLMs), we often grapple with the challenge of constraining their outputs to meet our specific needs. In this post, I will share a technique I developed to enforce structured responses from LLMs, parti...

3 min read · [engineering.fractional.ai](https://engineering.fractional.ai/taming-llm-responses-dynamic-pydantic-models-for-flexible-structured-output?utm_source=hashnode&utm_medium=feed)

[Aug 30, 2024](https://hashnode.com/posts/taming-llm-responses-dynamic-pydantic-models-for-flexible-structured-output/66d25a08fac976b81f12d073)