Skip to content

Show HN: PromptLayer – Track, debug, and maintain GPT prompts in prod

promptlayer.com
29 pointsjzone311 comments
On HN

Comments

Is this like Weights & Biases for prompt-based LLM experimentation (as opposed to model training)?

That's not a bad description. I don't have a good comparison— we mostly just built this as a tool that we personally needed. I like to think of it like W&B but also in the world of logging tools like Prometheus.

LLMs are a funky thing to work with, and we need a whole suite of new tools to make them prod-ready.

Sounds like it, pretty smart idea

How is it trusted by 400 engineers already?

400 people signed up maybe?

What is a "prompt engineer"?

https://www.google.com/search?q=prompt+engineer

Prompt engineering is a concept in artificial intelligence, particularly natural language processing (NLP). In prompt engineering, the description of the task is embedded in the input, e.g., as a question instead of it being implicitly given. Prompt engineering typically works by converting one or more tasks to a prompt-based dataset and training a language model with what has been called "prompt-based learning" or just "prompt learning".[1][2] Prompt engineering may work from a large "frozen" pretrained language model and where only the representation of the prompt is learned, with what has been called "prefix-tuning" or "prompt tuning".[3][4]

From https://en.wikipedia.org/wiki/Prompt_engineering

I asked ChatGPT for you:

A prompt engineer is a person who designs and creates user interface prompts, such as chatbot prompts or command line prompts, to communicate and interact with users effectively. They are responsible for defining the language, tone, and structure of the prompts in order to guide users and elicit the desired response.

For our purposes, anyone building on LLMs and tinkering with the prompts to use.

Someone who writes prompts for AIs. It's what's going to be left when all the programming jobs are gone.

A GPT whisperer

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.