Proud to share my submission for the Frontend Hackathon by Sheryians — an exclusive challenge for cohort students. I built a UI-focused website for Rosier Foods, the brand by Gaurav Taneja (Gaurav Taneja). live on : - rosier.vercel.app repo- github.com/Akki-b2008/Ros…

#Day13 of Backend Grind, Today we locked down our routes with JWT + Cookies – Issued tokens on register – Stored them in HTTP-only cookies – Used cookie-parser to read them – Protected /user with jwt.verify() #NodeJS #JWT #XSS #BackendDev

#Day14 of Backend Learning with Ankur Prajapati Started building a lite Instagram clone, With AI-generated captions too. Of course. ✅ Understood the backend flow ✅ Set up Express + MongoDB ✅ Created a lonely /register route Tomorrow: Auth madness begins. 🔐 #BackendDev #NodeJS

#Day15 of Backend Dev 🧠 Wrote /register, /login, /user, /logout 📦 Moved logic to controllers/ like a “real dev” 🔐 Hashed passwords with bcryptjs — because storing plain text is... bold Next up: AI-generated captions. Because why write them when the bot can flirt for you?

Ab to Seedhe google me

#Day16 & 17 learning backend with Ankur Prajapati -Built a /posts API: – Custom auth middleware that checks cookies 🍪   ⤷ No token? No entry. – Clean structure with separate controllers – Captions via Gemini AI 🤖 (it speaks more confidently than I do) #Backend #LearnInPublic

Day 18 – backend + project update with Ankur Prajapati bhaiya what i learn today – Rebuilt the whole project from scratch – Created /post route – Converted image to base64 – Sent it to Gemini AI for caption generation practice actually works. #NodeJS #backend #learnInPublic

#Day20 with Ankur Prajapati Bhaiya – How AI really works behind the scenes – Power of RAG (Retrieval-Augmented Generation) – Difference b/w Short-Term & Long-Term memory in AI – Deep dive into Socket.IO & WebSockets – real-time magic #Backend #SocketIO #WebSockets

#Day20 of learning backend with Ankur Prajapati Explored persistent WebSocket connections Built event-based communication using Socket.IO Understood real-time chat flow Gemini-powered AI chatbot now replies live Backend unlocked a new level 🚀 #NodeJS #WebSockets

#Day21 of backend with Ankur Prajapati bhaiya Project update, Built: ⚡ Real-time chat w/ Socket.IO 🤖 Gemini AI chatbot 🧠 Gave it short-term memory (because AI forgets... fast) Learned: AI isn’t magic — it’s just a lot of smart wiring 🛠️ #BackendDev #NodeJS

#Day22 with Ankur Prajapati bhaiya started a new project with ejs & express – SSR vs CSR: Who renders it better? – Set up EJS + Express for dynamic views – Peeked into multi-core processing (MCP 🤯) #100DaysOfCode #NodeJS #BackendDev

Ek hum the aur ek ye aajkl ke चिलगोज़ें

After a 5-day health break , I’m back stronger w/ Ankur Prajapati #day(24-25) of learning backend, – Rebuilt project from scratch – Register/Login + Auth middleware – EJS protected routes – Socket.IO logic moved to /socket folder Comebacks > setbacks 🚀 #neverGiveUP

#Day26 of backend grind w/ Ankur Prajapati bhaiya, Last night → doubt session turned into a deep dive on Redis 🔥 Now I get why it’s a must-have for our GPT project. Late nights, big brain mode. #BackendDev #Redis #GPT #100DaysOfCode

#Day27 of backend w/ Ankur Prajapati, Shifted our ChatGPT project from EJS → MERN. Why? MERN = full-stack power, React UI, scalable features. EJS was good for SSR, but MERN lets us go next level #MERN #FullStack #Backend #100DaysOfCode #AI

Abhi to start hi kia h bro,

#Day28 of learning backend w/ Ankur Prajapati bhaiya, Today, I started learning how to build that magic ✅Socket.IO for live chat ✅Middleware to check valid users ✅AI that remembers what you said before Not done yet, but backend suddenly feels alive. #AI #Backend

#Day29 of learning backend with Ankur Prajapati bhaiya, Today I realized AI’s memory isn’t magic — it’s math. Vectors + embeddings = the brain of AI 🧠 That’s how it remembers conversations, powers search, and makes recommendations. #AI #BackendJourney #LearningDaily

#day30 of backend w/ Ankur Prajapati Ever noticed AI replies like it “remembers” you? Truth: AI ≠ human brain. It has 🧠 STM (chat only) & 🧠 LTM (Vector DB). With RAG, it retrieves info → then generates answer. Your take 👉 Should AI have human-like memory or stay limited?