john-titor

Hi HN,

I am a complete LLM beginner but would like to get into the practical application of the technology by interfacing a GPT instance with an internal database at work.

I understand creating LLM agents that handle unstructured data is a quite common use-case. Ironically, more experienced peers also told me that going through the same exercise to make agents work with structured data can be more challenging.

Are there any resources that you could point me to wrt best practices and tools to use when tackling such a project? In my mind, I would 'magically feed' the DB schema to the LLM, have it write valid SQL prompts and translate the results into text. Does this make sense? Are there better ways to do this?

Any data scientists here that communicate their results or data sets through interactive plots? I was wondering what your favorite tool or package would be for this task. Working mainly in python, I often find myself using plotly or bokeh but find both packages' syntax clunky and unintuitive. Moreover, none of these scale well to large data sets - most of these solutions can not handle a simple 2d scatter plot with over 10k points. Commercial tools like Tableau are an option, but at the cost of some flexibility.

I love my job as a data scientist, it's incredibly creative and a lot of fun. But at the end of the day I'm pretty much done with staring at code snippets and reading documentations. The same is true for the weekends - ever since I started to code for work, my motivation to code outside of it rapidly fell towards absolute zero.

The problem is that I have a few private projects that I would love to try out. They all need coding. How do you find the motivation to work on such things when you also code for work?

Personally, I think it's probably a healthy thing to pursue completely diverse activities professionally and privately, but at the same time I feel like not honing one's professional skills during your time off is a little bit of a wasted opportunity.

I upvote HN posts for different reasons. Sometimes I just want to show appreciation for sharing a specific story. Sometimes it's a tool, or article that might be useful for {work, project A, project B}. Sometimes it's just funny random facts I would like to remember. Whatever the category, I was wondering if there is a good, easy way to organize the upvotes.

For my work I often have to present scientific topics to an audience. This includes a fair amount of notation in form of LaTeX math that has to go into the presentation slides.

So far, I have not found a good solution to efficiently include LaTeX in any presentation software.

There are plugins, but they rely on a local LaTeX distribution which, let's be honest, is always a pain to manage.

These days, I find myself using a browser app to render equations, then download and import the math as SVG or PNG. This somewhat works: The problem is that I have yet to find a web app that supports both, a wide range of commands (like https://katex.org/) or the export to SVG (like https://viereck.ch/latex-to-svg/).

Any suggestions how to optimize my workflow?