aymaneSennoussi

More context on how I did it: I used tesseractOCR to get the extraction and the layout analysis, then I passed the results as text to GPT4 api to possibly fix any misreadings based on completion. For example, if the text from OCR says "the couples were walking across the shiver", the LLM fixes it and makes it "across the river".

Hi HN,

I wanted to share something I’ve been working on over the past month and would love to hear your thoughts.

It’s a simple tool that converts PDFs to text while preserving layout elements like tables and headings, powered by LLMs.

You can try it with 500 free credits (no signup required) and get 1000 more if you sign up.

Would really appreciate any feedback especially from those who’ve struggled with messy PDF extractions before!

Thanks for checking it out.

I'm confused. This looks like a distillation of Qwen for math problems. What Am I missing?

Catastrophic! But with impunity as usual.

Skincare products.

You nailed it! I think Openai Team needs to patch it.

sure, but why would the JS not be scalable? Seems like a flaw on the frontend. The app is very simple and should not heat up an M3.

Any advice before I launch?

Hi, I built this tool that catches emails from unknown senders (Like salespeople) and labels them.

You can then either auto-archive them or just label them as outsiders.

This is something I couldn't do on Gmail, so I built this tool

Site: https://emailgurus.xyz

GitHub: https://github.com/asennoussi/emailgurus

There is practically no way to create filters on gmail based on whether the email is coming from your contact or not. So I built this open source solution that basically labels emails coming from people outside of your contact list.

PS. It's open source. Let me know your thoughts.

I was trying to figure out a way to classify emails coming from people outside of my contact list, but Gmail did not offer any solution. Emailgurus is here to solve exactly that.

I know that most people are privacy-aware and thus reluctant to give apps access to restricted scopes such as reading your emails and your contact list for the possibility of data & power misuse.

This is why I'm happy to announce a few things with this launch: 1- I don't care about your data. That's why I am encrypting all your sensitive information irreversibly. We are using SHA256 to encrypt your contact list, and all lookup is performed on Hash to Hash basis. 2- Other sensitive data such as Google Tokens are encrypted with secured rotational keys. 3- This software is open source! Anyone can verify points 1 & 2 and see how I process your data and ensure we're not misusing it. Here is the link to the repo for anyone wishing to explore/contribute. https://github.com/asennoussi/em... (Remember to it) 3- For everyone who doesn't want to bother with the code, the building, the deployment, the infrastructure, the DevOps, and just wants to keep their inbox clean and the noise to the minimum, you can use the service for a small monthly fee. (I'll be thankful for that.)

Please let me know your feedback and your comments. Join our Slack community for feature requests, discussions, etc.. (https://emailgurus.slack.com/)

Hello, this is stupid. Google doesn't penalize you for using GA, but because you're not using it the right way. If you want to use GA without JS overhead, use their API and build your own JS friendly tracker.