TheNeorav
50 posts


@code_bytein @neerajjj6785 What's the take here? If you are providing free service how are you getting paid? You get the Email lists that can be used for distribution or marketing?
English

@neerajjj6785 Adding some new features this weekend
mylinx.in
English

@HxrshitYadav Though I'd say currently non replaceable but the timeline is small.
English

@HxrshitYadav I think we might be underestimating here. Ai is already capable of replacing devs here. The big tech team needs more GPUs or they just need to find a way for efficient usage. If that's solved. Ai assisted coding would become the standard. Compare the past few years and you'll see
English

Nobody told me any of this in college.
I figured most of this out after wasting 3 years.
So if you are in a tier 3 college right now with no one to guide you,
this one is for you.
> How to actually win coming from a tier 3 college.
> No paid courses.
> Just an honest roadmap.
Pick ONE language first.
> C++ or Java or Python.
> Don't switch.
Master the basics before anything else.
> OOP
> Memory management
> Time and space complexity
If your fundamentals are weak, everything else will crack under pressure.
> DSA every single day.
Striver's A2Z sheet
Free
Perfect order
> Git and GitHub from day one.
This is your proof of work.
Recruiters will check this before they check your resume.
> Web basics next.
HTML, CSS, JavaScript.
Chai aur Code is genuinely good for this.
> Build one simple website.
Not to impress anyone.
Just to feel how everything connects.
> Then pick a frontend framework.
React or Next.js. Harkirat Singh or Chai aur Code. Either works.
> Backend.
Pick one stack and go deep.
Option A: Node.js + Express + MongoDB or PostgreSQL
Option B: Java + Spring Boot + MySQL
Do not try both. Pick one and actually finish it.
> Learn REST APIs properly.
Postman. Real projects. Not just theory.
> Build 2 to 3 full stack projects that actually solve something.
User auth. CRUD. A real world problem like a chat app, blog or e-commerce.
These projects will do more for you than your degree ever will.
> AWS basics on the free tier.
Deploy on EC2 or Vercel. Use S3. Learn IAM basics.
A deployed project beats 100 projects sitting only on your laptop.
> Before interviews, cover these four things properly.
DBMS, OS, Computer Networks, and System Design basics.
Not surface level. Actually understand them.
> Make a clean portfolio.
> Then apply everywhere.
Naukri, Instahyre, Company career portals, etc
Everything I just listed is free.
6 to 12 months of serious work and you are genuinely competitive.
Tier 3 college is not your ceiling.
Your consistency is.
Save this. Send it to someone who needs it.

English

@HxrshitYadav We will have it soon enough. Sarvam is doing their best. While a start up in Banglore is already running quantum computing. India is not the front runner right now but we will catch up.
English

USA has:
> ChatGPT
> Claude
> Gemini
> Grok
China has:
> DeepSeek
> Qwen
> Kimi
> MiniMax
Both countries are racing to build the future of AI.
Meanwhile, India has:
> 1.4B people
> The world's largest developer population growth
> Endless AI users
But where are our foundation models?
> Where is our DeepSeek?
> Where is our Qwen?
> Where is our ChatGPT moment?
The AI race isn't company vs company anymore.
It's country vs country.
And the question is:
What is India doing? 👀🇮🇳
English

@YashManghnani1 If it's not just tech and has purpose bleeding too much into reality and is too ingrained inside Businesses and irreplaceable by nature.
1. Shouldnt be just a software should be bound with reality based service
2. Irreplaceable quality and offer and service
English

@GohilHardy For code and more depth related things anthropic open ai for regular go to things. Gemini is only for use inside the Google infrastructure and may use everything for where it sounds to suitable your workflow
English

Right now, ai helps with generating codes but I noticed it's not causally aware of what happens in each layer when connected. I have made model where the full code base you built is read as world model and you'll actually understand the casual chain and what causes bugs. #ai
English

Grok recently compared Becomer vs other memory tools unprompted.
Verdict leaned toward Becomer for multi-tenant efficiency & cost.
Building in public is wild 😅
becomer.net
English

Most agent memory solutions are just "bigger context windows" in disguise.
Becomer.net is actual external persistent memory.
Store. Recall. Zero tokens wasted.
Any model. Any framework.
becomer.net
English

Building agents that actually remember and improve over time — without bankrupting you on tokens.
That's why I created Becomer.net: persistent external memory for any LLM.
Zero tokens for store/recall. ~100ms latency. Works across GPT, Claude, Llama, etc.
English









