EmaDev

@did0f @marclou 😂

@emadev01 @marclou ahaha we all are very similar on this I see

Am I overfitting the results? Wdyt? @marclou, can I study your X profile? 😅

@sebaa_dev @did0f 🙏🏻 I will keep you all updated

@emadev01 @did0f So interesting mate

It seems like @did0f is among the people I follow who share the highest-quality posts 🙌🏻

@did0f @marclou Me too 😆 that’s the problem, I need to stick to just a few products at a time

@emadev01 @marclou Makes sense. I like this project btw

@kail_designs Really good backgrounds 😁 I need some of them!

One of the best websites I designed recently✨ backgrounds.supply

@did0f @marclou Some predictions are too close to the actual stats 😅 I probably need a larger dataset

@emadev01 @marclou Why do think it may be overfitting?

@did0f @JayNaiduX I’m not really in my comfort zone when it comes to writing 😆 but I could share the repo with you privately 😅 (it might break some X T&Cs, so I can’t make it public). If you like it, feel free to contribute or write about it haha

@emadev01 @JayNaiduX Well I’d love to see that under the hood. Do you think you’ll do some tech articles at some point?

@JayNaiduX @did0f Actually it's quite raw, collecting feeds and stats, then calculating some few kpis, then trained a ml model to predict results

@emadev01 @did0f This is great man! I had this idea few days ago, but don't know how to execute it! But you did a great job. Curious how u build it! 😊

@did0f Nope, it’s local-only and internal for now. I built my own small ML model that predicts performance from stats, and I train it on a dataset I collect from my own X usage to keep improving it 😅

@emadev01 This was unexpected, a good kind of unexpected. Thank you! What is this dashboard? Is it publicly accessible?

@did0f Your stats are saying that 😂 It’s an internal tool I’m building to fine-tune my content based on similar profiles that perform well, and it turns out your profile has the highest quality in my circle 👏🏻

WOAH! What is this? And why are my posts at first place in quality?!?

I've been working on a small PoC that might surface some pretty interesting insights. First ever ML model trained locally to preview tweet engagement 🤯 No cloud, pure privacy #ml #llm #ai

Note: this post was improved based on the first PoC 🤯

Been deep diving into X audience metrics for ~3 hours. I’m thinking of building a local-only utility that studies your followers and suggests variants that actually convert, inspired by who YOU follow. No cloud, pure privacy. Wdyt? 🧐

@did0f Amazing ui, great sneak peek 😁 cool icon too 💪🏻

Dual captions coming the Velocaption V3. Stay tuned.

I have to say, it’s one of the few tools I’ve found that’s genuinely useful for everyday work. It’s saved me a huge amount of time on summaries, brainstorming, and more. I’m really glad my company has adopted it 🙏🏻 It’s helped me stay focused on higher-value tasks while letting Granola handle the “memory” side of things. 💣

Has anyone ever seriously looked into what behaviors the X algorithm actually rewards? I’m working on a small PoC that might surface some pretty interesting insights 👀

@ieeeedan Finally the theme editor also for the svelte fork! 💪🏻

shadcn-svelte/create landed today btw

you can fork the project here! f devmotion.app/editor/p/ti49n…

A ton of updates for DevMotion 🚀 - 💨 Rendering time reduced by 50% - 🎥 Added new filters, including a glitch video effect - ⚒ Introduced templates to get similar results in just a few clicks - 🤯 Almost finalized a solid solution for handling large files without cloud storage issues And don’t forget: you can tune your project with the LLM you prefer. Check out the project in the comments 👇