RECENT ROTATIONS

In this post, we explore sending a secure message to a user that doesn’t have an account on your system and therefore cannot be authenticated. We’ll explore the use of HMAC verification tokens that you can email to a recipient …

I spent last week in San Francisco meeting up with the Rotational team to attend PyTorch Conference. If you’re an LLM developer and didn’t make it this year, here are some of my key highlights and takeaways.

If you’ve worked with generative AI models you know they can be fickle and sometimes fail to meet the expectations of users. How can we move towards models users trust and see clear value in? Let’s engineer a user-feedback loop!

Anyone who has worked on a POSIX system is used to fine-grained data access rights for files and directories. In a world of generative AI, do we need a new permissions model for files used to train LLMs and serve live inferences?

Messy rooms are a hassle. They can leave you sifting through piles of random things until you find what you’re looking for. In the messy world in which we live, the progress of the recommender system and the many approaches one can take in …

Why is everyone so excited by AI? Chatting with a bot isn’t the game changer. Instead, AI represents a new method of interacting with complex computing resources and data – one that is far more accessible than previous methods.

Over the past few years Atlanta’s tech scene has grown at a rapid pace. There was no better place for proof of this than at RenderATL 2024.

winkJS is a collection of open source libraries, created by Graype Systems, designed to make Natural Language Processing (NLP), machine learning, and statistical analysis accessible in JavaScript. Let’s take an in-depth look into …

As industry races for use cases of Large Language Models, software devs have emerged as early adopters. Can LLMs help us translate between tech and talk? Let’s build a text-to-SQL application with Vanna and Streamlit!

Sick of hearing about hyped up AI solutions that sound like hot air? 🧐 Let’s use boring old ML to detect hype in AI marketing text and see why starting with a simple ML approach is still your best bet 90% of the time.