Ahmed Elhadidy

Egyptian 🇪🇬 MMPME - MUH - class 2026, researcher, entrepreneur, software developer & data nerd, USMLE instructor & tutor, Founder of TEMP-U.

[Ahmed Elhadidy @A_Hadidy_]

Alhamdulellah

USMLE step 1 ✅
USMLE step 2 ✅
One remaining ⏱️

[Ahmed Elhadidy @A_Hadidy_]

Sorry my friends, it hit a rate limit

I will try to put a new link tonight

[Ahmed Elhadidy @A_Hadidy_]

I have decided to giveaway my USMLE Step 2/3 Anki for free for everyone

This helped me to score a 275 2 years ago

It’s a little bit big, but i’m sure it’s of very high quality

I wish you all success

https://t.co/BUseVUkiqh

[Ahmed Elhadidy @A_Hadidy_]

Sorry but if the link is not working it’s because of rate limit because of the number of downloads

So in a few hours try again

[Ahmed Elhadidy @A_Hadidy_]

I have decided to giveaway my USMLE Step 2/3 Anki for free for everyone

This helped me to score a 275 2 years ago

It’s a little bit big, but i’m sure it’s of very high quality

I wish you all success

https://t.co/BUseVUkiqh

[Ahmed Elhadidy @A_Hadidy_]

@yashu211996 @SARA__AMER_ Sure, no problem

[Ahmed Elhadidy @A_Hadidy_]

@WajahatAli87 Yes, very possible, download ankidroid

[Ahmed Elhadidy @A_Hadidy_]

@WajahatAli87 Hi, yes it’s safe, i have uploaded it directly from my laptop

It’s made of 8200 cards, that’s why it is big

Install Anki app on you laptop

From inside the app you’ll have an “import” button

Import that file, it’s basically a database of flashcards

https://t.co/mjhHmsPZRF

[Ahmed Elhadidy @A_Hadidy_]

@AbdelmunemAli @Ahmed_ElHadidy_ العفو يا حبيبى، بالتوفيق يارب

[Ahmed Elhadidy @A_Hadidy_]

@_mooo10 آمين، الله يسلمك

[Ahmed Elhadidy @A_Hadidy_]

@Ghalwash8 You’re welcome 🙏🏻

[Ahmed Elhadidy @A_Hadidy_]

@Leon_SKC شكراً، آمين

[Ahmed Elhadidy @A_Hadidy_]

My dear friend @moibra359 also decided to giveaway his USMLE Step 1 Anki deck

https://t.co/uiL3QIfRVE

[Ahmed Elhadidy @A_Hadidy_]

In medicine, that could mean:

→ Q&A over patient records
→ Searching semantically through huge medical datasets
→ Q&A over recent clinical guidelines
→ Connecting AI to your own medical knowledge base

RAG is basically giving an LLM the right memory at the right time.

[Ahmed Elhadidy @A_Hadidy_]

When talking about AI and Large Language Models in medicine, there’s one concept you need to know:

RAG — Retrieval-Augmented Generation.

[Ahmed Elhadidy @A_Hadidy_]

The cool part is how it works:

Imagine a multidimensional universe of information, where things with similar meanings live close together.

Ask a question → find the closest information → give it to the model → generate the answer.

[Ahmed Elhadidy @A_Hadidy_]

The third publication in 5 months.

الحمدلله

You can learn more at: https://t.co/7e9EsCUgys

[Ahmed Elhadidy @A_Hadidy_]

@akashcorex I keep the task very small and hallucination almost never happens, so one site per task, i collect and relate the data in sqlite, then pass over it again to scrape

It’s a chain architecture one gets and saves

[Ahmed Elhadidy @A_Hadidy_]

How to scrape/crawl a website in a few steps in late 2026, without knowing how to code, just copy and paste these steps to Claude or Codex.

1. You need links to all the relevant pages on that website, for me i have an agent for that, you can do it manually.

[Ahmed Elhadidy @A_Hadidy_]

Hello 👋🏻

I have programmatically crawled the web to create this Notion database for NRMP Match Programs in General Surgery & Pediatrics, mainly with data from 2026 cycle, rarely from 2025, i hope it’s helpful.

The url is in the first reply.

[Ahmed Elhadidy @A_Hadidy_]

@DaliaMISelim You’re welcome

[Ahmed Elhadidy @A_Hadidy_]

Hello 👋🏻

I have programmatically crawled the web to create this Notion database for NRMP Match Programs in General Surgery & Pediatrics, mainly with data from 2026 cycle, rarely from 2025, i hope it’s helpful.

The url is in the first reply.

[Ahmed Elhadidy @A_Hadidy_]

6. Save the results in any shape you like (eg, obsidian markdown files).

You can really customize this into something more powerful if you want, i wanted to keep it simple.

Finally please respect the scraping policy for those websites.

[Ahmed Elhadidy @A_Hadidy_]

5. Finally, convert that structured JSON into a markdown template again fed with the actual data, programmatically.