Backend-Focused Full-Stack Developer | GenAI

You don't have a forgetting problem. You have a remembering problem. Introducing DailyBrief — an AI memory assistant that works through WhatsApp. Send messages naturally. It remembers what matters and brings it back when you need it.
3
1
22
761
Grateful to have attended the AI Agent Masterclass by @surajtwt_ Happy to finish Top 2 out of 263 participants in the quiz Learned more about LangGraph, agent workflows tools and state Thankful for the opportunity to learn and test my understanding🙌 @Hiteshdotcom @nirudhuuu
3
1
30
400
A few months ago, I joined a GenAI cohort to understand how AI applications actually work. Today, I completed it. Along the way, I built RAG systems, AI agents, MCP servers, voice agents,and developer tools. But the biggest shift was learning to ask: What happens underneath?
4
1
28
940
I started looking beyond the LLM call. How do we design the backend? Handle tools, workflows, memory, and failures? Still a lot to learn, but my direction is clearer. Grateful to @Hiteshdotcom @piyushgarg_dev @nirudhuuu . Certificate earned. The real building starts now. 🎓
1
13
121
24.51 billion UPI transactions in August 2026. That's roughly 9,150 transactions every second sustained, for 31 days straight. As a backend dev, one question stuck with me: what happens in the second between scanning a QR and seeing "Payment Successful"?
1
12
276
Now put that at scale. ~9,150 transactions/second, averaged across August, means failures and delayed responses are not theoretical edge cases. Networks fail. Banks timeout. Responses get delayed. At this volume, failure handling isn't optional.
1
2
135
To Tejas: Scan → Authenticate → Pay → ✓ To the systems underneath: Routing → Authorization → Debit → Credit → Confirmation → Failure handling → Reconciliation 24.51 billion transactions look simple from the outside. The engineering underneath is anything but.
3
33
Today was the farewell of our GenAI cohort. It feels a little strange seeing this journey come to an end. What started as a learning opportunity became months of building, experimenting, making mistakes, debugging, and growing together. @Hiteshdotcom @piyushgarg_dev @nirudhuuu
1
25
435
I’m genuinely grateful to @nirudhuuu sir for sponsoring this opportunity and making this journey possible. And thankful to @surajtwt_ @BlazeisCoding @devwithjay @yntpdotme everyone in the cohort for the learning and memories. One chapter ends. The journey continues. ❤️
5
62
I made a brand-new Instagram account. Barely touched it. It still recommended school friends, college friends, my brother, my sister and someone whose number I deleted 3 years ago. So I dug into how Instagram's Suggested Accounts system actually works.
1
39
2,395
Important limit, stated plainly: Meta's docs explain the categories of signals and the pipeline shape. They don't publish the exact feature values or score behind any single recommendation. So I can explain the mechanisms. I can't tell you which one fired for me.
1
2
124
The real takeaway isn't "Instagram knows your friends." It's that a recommender doesn't need one clean signal. It can combine many weak, indirect ones graph, contacts, behavior into a single confident-looking prediction. That's the actual engineering story here.
2
107