Leveraging AI and Building Daily | Software Dev
Joined September 2022
- Tweets3.6K
- Following963
- Followers269
- Likes5.6K
pandeyji retweeted
the challenges if you're curious :)
fly.io/dist-sys/1/
pandeyji retweeted
WEB3 Builders is back.
Space 11: @prashant__sha hosts @kirat_tw.
Everyone asks Harkirat how to land the remote job. We ask what comes after: how you get paid from India, in dollars and in crypto.
Thu 15 Oct, 8 PM IST. Set a reminder:
xitter.cf/i/spaces/1qJDzWpoaNkKV
Best place to get info about government exams -
governmentexams.wiki in newspaper style.
Free tools + verified exam dates, fees, and eligibility - no login required.
Saare sarkari exam sab ek jagah.
Govt exams are now easy to view and follow.
Suddenly all possible posts on my feed is about how to become inference engineer from everyone
Did jev also started producing inference engineers😭😭
Was looking for this sort of editor!
Last 24 hours have been surreal; I never imagined this kind of response. What I am happier about is the feedback :) Acting on it, I shipped two of the most requested features
> px0 is now git aware and can render markdown files
so, from v0.1.2, you would see...
- files that are added or modified are marked
- git diff split/unified view
- filtering files that are added or modified
- preview for markdown files
If you have already installed px0, run the following command to update:
> sudo px0 --update
If you have not tried px0.ai yet, give it a shot ... it is a fast, simple, lightweight, browser-based, read-only IDE ... hyper-optimizing for performance.
loads in less than 1s and takes ~20MB of RAM.
pandeyji retweeted
Anthropic engineer:
“90% of our engineers were using self‑improving loops. Now everyone shifted to building agentic Graphs"
“No more prompting.”
In just 10 minutes, she builds her entire Claude Code setup and workflow live from a blank terminal.
This is more valuable than most $1000 agentic courses.
Watch this video, Then save the article below if you want to become a graph architect before everyone else catches up.
Just built something for fun :
parlour.wtf
Do check it out :)
A sister site to saloon .wtf out : parlour.wtf
One for the women
pandeyji retweeted
I’m the actual human who made every single one of these videos they keep gluing together without asking. They love calling me ‘ex-Google engineer’ or ‘this guy’. Follow this guy and let’s actually chat. Original post is right here:
xitter.cf/shenseanchen/status/20…
Want the real AI Agent Harness System (Loop + Memory + Eval) instead of the mash-up? Grab my open-source Waku Agent (almost 1k stars): GitHub.com/shenseanchen/waku…
Original YouTube channel: yt.seanchen.io
(Better for your eyes. Trust me:))))
I also run a community where I host live biweekly Q&As and share all the original system-design files:
whop.com/sean-ai-community-d…
@X @elonmusk @benjitaylor @singhai @dinkin_flickaa please fix the theft on X.
@vattybear thanks so much for the constant support my friend! 🔥
THIS GUY MAKES COMPLEX AI AGENT CONCEPTS RIDICULOUSLY EASY TO UNDERSTAND.
No jargon wall. No assuming you already know what a reward model or an evaluation harness is.
Just a straight line from "I have no idea how agents actually work" to "oh, that's genuinely simple."
The best explainers do not simplify by cutting corners. They simplify by finding the one analogy that makes the whole thing click.
pandeyji retweeted
best way to learn docker is by using it. Use Docker for everything you are running right now, unless there is a really specific reason why that is impossible.
The clue for me of really understanding the use case, was being confronted in production with configurations that could take hours or generate very complex automation setups. Dockerizing the setup is so powerful in that use case that you learn how to use it and what it's strong points are right away.
If you do not have a complex production setup or are just learning on your own, try to create something like that. Set up an elasticsearch cluster with kibana, set up a prometheus / grafana, set up some SQL databases, ... Just things that you would do in real life. You will be confronted with everything you need to run a complex setup in production and you will learn really quickly how much more easy the Dockerized setup usually is vs the native.
pandeyji retweeted
Every backend engineer should know how distributed systems work and their basics.
If you want to actually master Distributed Systems, stop hoarding tutorials and watch Martin Kleppmann’s free lecture series.
It completely changes how you think about building scalable, fault-tolerant backends.
Link in the replies. 👇
pandeyji retweeted
yoo chat
if you are looking for any repositories to dive deep on the topic of docker image optimization than have a look at this one .... hope it helps you
github.com/ag-chirag/ai-dock…
So I was building an ambitious project which I will talk about sometime later here on X, and thought of sharing all sorts of free ways to use AI as long as you can :
1. ChatGPT Go for high quality prompts and brain storming (got this for free previous year) ,
2. Google AI Studio Provider API keys used directly in VS Code (rate limit reached quickly as the project grew) (got gemini pro free via jio) (but not that high rate limits)
3. Used Kilo Code VS code extension with its free models from nvidia (from 2nd prompt onwards it started giving hallucinated garbage output)
4. Switched to blackbox ai vs code extension (Kimi K 2.7 was provided free with it -it tried hard but yet again reached its limits in 2nd prompt)
5. Could not use Kimi Code as monthly quota over, also its very slow on free tier
6. Mini Max Code was also glitchy on Windows (wasn't even opening after closed once), hence not used
7. Z code required subscription so automatically out of this list
8. Avoided Qwen CLI (looking at the locations they store data). I avoided Qwen Code too (its same as qwen chat ui interface , idk why there's the suffix 'code' to it)
9. Tried COMPOSER 2.5 from Cursor (limit reached in first prompt as project is big)
10. Come to rescue Open Code (utilized it with CloudFlare Workers AI API keys, limit reached then switched to Open Code Zen's Big Pickle - maybe its GLM-5.2 ? ) Big Pickle was the ultimate saviour after all the rate limits
What if all the above got exhausted ?
I still have :
- Codex coding agent
- continue. dev coding agent
- Nvidia api keys (from build . nvidia. com)
- Grok cli
- Anti gravity's own limits.
I should have integrated and worked with more 'skills' into my projects like hand-off .
What's stopping you from coding today ?
pandeyji retweeted
ended up finding this book on the matter. it's apparently a more academic, practical version of DDIA.
and it's free!
