Atharv Sathe

Software Engineer | Web Developer

[Atharv Sathe @AthSat7]

Hi folks 👋!

I'm Atharv Sathe, an aspiring Software Developer and persuing a bachelor's in Computer Science and Engineering from VIT Bhopal University.

🌐 I am currently doing full-stack Web Development, creating seamless user experiences and robust back-end systems.

[Atharv Sathe @AthSat7]

Built my first CI/CD pipeline for home server (StarGate).
Original App Size: 1.1GB
Docker Image: 650MB (>45% Reduction)
CI/CD Pipeline Time: 1m 45sec
Took more than 15 failed attempts to get the first successful deployment.

[Atharv Sathe @AthSat7]

I'd love your feedback!
⭐ Follow me on GitHub : https://t.co/a4nVNc9H0T
🐞 Bugs or feature ideas? Drop me a line at [email protected].
#ZeroSum #Fintech #AI #NextJS #LangChain #LangGraph #BudgetTracker #PersonalFinance #ProjectLaunch

[Atharv Sathe @AthSat7]

Introducing ZeroSum v1.0 🚀

A minimalist, AI-powered budget tracker designed to bring clarity to your personal finances.

Take control of your money, the simple way.
Check it out live! 👇
https://t.co/cU2Gcl9X7E

[Atharv Sathe @AthSat7]

The mission isn't over. Here's what's coming next:
🧠 Enhanced AI for Kowalski (Ask any financial question!)
🔒 End-to-End Encryption for your data
📧 Automated Email Reports delivered to your inbox
Stay tuned for more superpowers! ✨

[Atharv Sathe @AthSat7]

@GithubProjects https://t.co/a4nVNc9H0T

[Atharv Sathe @AthSat7]

https://t.co/j5WxT6sakz

[Atharv Sathe @AthSat7]

The night is dark and full of... bugs.
Winter is Coming 🐲
#winteriscoming #winterarc #buildinpublic

[Atharv Sathe @AthSat7]

Can highly relate to "Learning on Demand" rather than mugging everything just because it may come handy later on...

[Ruben Hassid @rubenhassid]

Andrej Karpathy says you should learn AI depthwise, not breadthwise.

Most education is breadthwise: watch lectures, memorize formulas, and trust you'll need it later.

Karpathy flips this by learning "depthwise, on demand."

What this means:

Pick a project, start building, and learn exactly when you hit a wall.

When he created a tutorial on transformers (the architecture behind ChatGPT), he didn't start by explaining attention mechanisms or complex architectures.

Instead, he started with the simplest possible thing: a lookup table that predicts the next word.

You build that first. Then you try to make it handle more complex patterns.

And it breaks.

Only then, when you've felt the limitation, does he introduce the next concept. Each piece solves a problem you've actually encountered.

As he puts it:

"It's a dick move to present the solution before I give you a shot to try it yourself."

When you attempt the problem first, the solution actually makes sense.

Teaching forces you to learn. "If I don't really understand something, I can't explain it."

When you try to explain and stumble, you've found the gaps in your understanding.

...

Build a project that gives you a reward.

Hit a wall. Learn just enough to solve it. Then explain it to someone else.

Don't consume content. Build the code.

That's how you actually learn.

[Atharv Sathe @AthSat7]

Middleware in NextJS when deployed on Vercel is designed to run on Vercel's Edge Network while actual server functions are executed in centralized regions.
I sort of had a faint idea about this but this screenshot of vercel's logs makes it even more clear...

[Atharv Sathe @AthSat7]

ummm... don't know how important this is... 😅

[Rohan Paul @rohanpaul_ai]

This paper introduces a new method called Agentic Context Engineering (ACE).

It helps language models improve by updating what they read and remember, instead of changing their core weights.

Normal methods that edit prompts tend to make them too short and lose important details over time.

ACE fixes this by treating the model’s context like a growing notebook that keeps and organizes useful strategies.

It has three parts: a Generator that works on tasks, a Reflector that learns from mistakes, and a Curator that updates the notebook with helpful lessons.

Instead of rewriting everything each time, ACE makes small, local edits, which keeps learning fast and preserves past knowledge.

It also trims repeated or useless points so the context stays clean and focused.

In tests with coding agents and financial reasoning, ACE gave about 10% higher accuracy and cut learning time by about 87%.

It even worked well without labeled training data, using only real-time feedback from how tasks ran.

----

Paper – arxiv. org/abs/2510.04618

Paper Title: "Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models"

[Atharv Sathe @AthSat7]

Gotcha !!
SQL 50 🔥🤩😁

[Atharv Sathe @AthSat7]

Finally back at DSA .... after a long break.
Crossed 500 on LC... 🔥

[Atharv Sathe @AthSat7]

Learnings :
1. Don't make a fantasy list of goals assuming you are the most disciplined, focused and determined guy on the planet.
2. Time Once Wasted is Forever Wasted... thinking that I will pull 2 - 3 all nighters to make up for it is wishful thinking.

[Atharv Sathe @AthSat7]

Starting this thread to track my progress for #15DaysOfDiscipline
Goals :
1. LC Daily
2. AtCoder DP Contest
3. DBMS (Theory) + SQL (45 Qs)
4. Complete Echoes
5. AWS Cloud Practitioner FCC YouTube

less... go !! 🔥🔥

[Atharv Sathe @AthSat7]

The first #15DaysOfDescipline challenge is over ... not a great result but learnt many things...
Results :
1. LC Daily (12/15) : 80 %
2. AT Coder DP (4/16) : 25 %
3. DBMS Theory + SQL (45 Q) : 0 %
4. Complete Echoes : 25 %
5. AWS Cloud Prac : 0 %

[Atharv Sathe @AthSat7]

Is it just me who feels the need to know the total duration of a you tube playlist ... ? nevertheless...
Created a simple basic extension to calculate the total duration of a you tube playlist. Contribute and suggest new features....
https://t.co/a75GLzxGgj