martimchaves

Hey y'all, couple of projects to show off:

https://ragbandit.com - improve the retrieval stage of your RAG systems by tuning your document processing pipeline

https://smolinvoiceagent - an agent that process invoices, you make corrections, the agent learns your ways

https://vendor-simple-central.streamlit.app/ - this is just a POC, but it's a system to process and extract insights from data from amazon's vendor central

This post is really wholesome :)

This is really cool, any challenges in developing this? How did you decide on the directions that you take when writing a word?

Couldn't get the letter constellations working on my end.

Country quizzes is a weak spot of mine, loved that. Would be cool to move the globe! Also, kudos for the bus cataloging!

That sounds really cool, how are you going about this? Are you training a model using your own EEGs as inputs?

I love tiled words, I've actually been doing the daily puzzle for a while, have completed over a 100 of them! It's part of my morning routine :)

I'm not sure if it would fit the theme, but sometimes I end up searching what an expression means, or where does it come from. Maybe it would be cool to have a little info box after you discover what the word is. Just an idea! Not sure if it would clutter things, and you can always search it yourself, but something I've been thinking about. I still remember looking up peanut gallery and sand dollar!

Very fun game, thanks. When you try to change the speed of the game on mobile using firefox, it glitches and keeps opening the speed modal. Pretty much crashes the browser. Also +1 to the taxi fare not being paid in the currency of the country, but in some other currency. More jobs would be great! Appreciate the difficulty of not knowing exactly where you'll be taken, but without consistent jobs sometimes it's really punishing (or at least feels really punishing). I was thinking it might be cool to add stamina, and fancier hotels may replenish more stamina. An incentive to splurge ah.

I'm guessing that human accuracy may be lower or around that value, given that handwritten notes are generally difficult to read. A better metric for document parsing might be accuracy relative to human performance (how much better the LLM performs compared to a human).

Cool read! I loved it when you changed the numbers to base 15, I thought that was a beautiful solution.

I feel you, exact same routine here.