I genuinely enjoy development—not because I’m chasing a trendy job, aiming for money, or trying to compete with those who grind DSA or competitive programming around the clock. It’s simply something I find fulfilling on its own. Be own your own

We're nearing 2026 with the same dull tech hiring practices. Consider this: > DSA is largely irrelevant. Companies face architecture issues, not syntax/code optimization problems. > 3-4 interview rounds are pointless. > Focus on system design and problem-solving tasks. >

Got AI Engineer Internship in a Banglore Based Startup📈 Will be cracked for a few months!!

Anyone in Banglore near Hustle Hub Techpark or HSR Layout??? Need Boys PG

Not selected , due to lack of core knowledge related to transformers and vector DBs

Round 1 is scheduled for Gen AI intern 🚨ZOMATO Anyone with an R1 interview experience in Gen AI ???

Everyone talks about Agentic & Generative AI, but few see the real work: optimizing pipelines, training models, handling hallucinations, and scaling for production. Flashy AI is cool, but production-ready AI is engineered. 🚀

Ever built something on your laptop that suddenly becomes alive? 😅 Say hi to OlaAssist: Whisper + Pipecat + LLM + FastAPI = AI call agent that talks, thinks, and maybe replaces your call center 😜 Next stop: Concept Usage Final Year Project detecting Early Parkinson’s via voice

2035: -Ethereum -Solana -DApps -DeFi -BaaS

Interesting update: Deepseek is the boss🫡; none of us have used or heard it repeatedly and solely. Results are from @dhruv_rathee 's @aifiesta

Citations: 𝑮𝒂𝒓𝒕𝒏𝒆𝒓, 𝑰𝒏𝒄. (2022). 𝑭𝒐𝒓𝒆𝒄𝒂𝒔𝒕: 𝑳𝒐𝒘-𝑪𝒐𝒅𝒆 𝑫𝒆𝒗𝒆𝒍𝒐𝒑𝒎𝒆𝒏𝒕 𝑻𝒆𝒄𝒉𝒏𝒐𝒍𝒐𝒈𝒊𝒆𝒔 𝑾𝒐𝒓𝒍𝒅𝒘𝒊𝒅𝒆, 2021-2026. 𝑮𝒂𝒓𝒕𝒏𝒆𝒓 𝑹𝒆𝒔𝒆𝒂𝒓𝒄𝒉. 𝑴𝒂𝒓𝒄𝒖𝒔, 𝑮., & 𝑫𝒂𝒗𝒊𝒔, 𝑬. (2019). 𝑹𝒆𝒃𝒐𝒐𝒕𝒊𝒏𝒈 𝑨𝑰:

Agents do not replace core ML. They are as bad as the data and models on which they are based. In the absence of good modeling, data science, and feature engineering, you simply drag-and-drop blocks and hope that the black box does not give you a wrong answer.

There is a boom in agentic, no-code products such as n8n [stitching workflows without ML code writing]. I have created one myself: A reddit post automating system. It was fast to deploy, and it showed something significant.

What the heck!!!!

I don't want it to be fast. I want it not to hallucinate. @OpenAI

System uptime: 3:10:23:51 on my 12th Gen Intel, 15GB RAM setup. Says All🫩

Phone - 24k earbuds - 1.2k earphones - ❌ watch - 2.5k laptop - 56k tablet - ❌ now your turn ??

@PratapSarnaik @mohol_murlidhar @ChDadaPatil @MLABansodeAnna @mieknathshinde पुण्यातील पीएमपीएमएल बसेसमध्ये काय बिघाड होत आहे? कंडक्टर वेगळ्या ठिकाणी (बस जाते, प्रवासी जिथे चढला होता तिथून नाही) आणि विचारल्यापेक्षा जास्त अंतरावर जाऊन जास्त पैसे आकारत आहेत ! ती सुरक्षिततेची चिंता असू शकते!

Tagging all the concerned authorities. @PratapSarnaik @mohol_murlidhar @ChDadaPatil @MLABansodeAnna @mieknathshinde

What’s going wrong with Pune’s PMPML buses? Conductors are overcharging by entering a different origin (that the bus passes, not where the passenger boarded) and a farther destination than asked. Can anyone explain this? also it is a security concern. @PMPMLPune @PMCPune

Grok faces ML challenges in verification, including noisy X data impacting LLM accuracy, limited deepfake detection due to lacking specialized CNNs, highly rapid response for real-time processing, contextual misinterpretation from biased training data.