Simran Tech

@X 167 verified followers. Road to 500 starts now. Developers, founders, indie hackers, and AI builders: Drop your hi or project below 👇

@yadavji_codes Fair. What industry looks more attractive to you right now?

@devfrom_hyd I would just leave this tech world

If you were 18 again in 2026, what would you spend the next 5 years mastering? No memes. Serious answers only.

@bytesizedshrey @X Sure

@devfrom_hyd @X let’s connect

Hey everyone 👋 Looking to connect with verified followers only on @X. Interested in: • AI 🤖 • Tech 💻 • Coding 👨‍💻 • SaaS 🚀 • Startups 📈 If you're verified and active in these spaces: ✅ Follow me — I'll follow back ✅ Turn on notifications — I'll do the same ✅ Engage with my posts — I'll engage with yours Building a genuine network of creators, builders, and professionals. Verified accounts only 💙 Drop a 👋 below and let's connect.

NEVER HIRE A VIBE CODER AGAIN !!!

@ravikiran_dev7 lol

Entire world: We need more GPUs Meanwhile, Nvidia CEO Jensen Huang:

@Its_Nova1012 Good pick. Would you focus more on the theory or building real-world applications?

@devfrom_hyd AI/ML and cybersec are future

@md_kasif_uddin yes

@devfrom_hyd Never tried it, is it good?

Be honest, which auth do you actually use in production?

@timeout_7 yeah

@devfrom_hyd Yes have you visited ?

Guess the metro Station 👀?

@md_kasif_uddin so what are you planning?

@devfrom_hyd I am 18 🥲

@sebuzdugan Interesting shift. AI labs are moving from chatbots to critical infrastructure. Biotech might be one of the biggest AI stories of the next decade.

OpenAI just opened GPT-Rosalind to US biodefense agencies, and launched a sponsorship program for life-sciences AI startups Initial partners: JHU Applied Physics Lab and CEPI

@adahstwt macOS

in your opinion, which OS is best for non-tech person?? -macOS -windows -Linux

@VikasAlwys AI/ML seems to be a recurring answer here 😄 What area of it interests you the most?

@devfrom_hyd AI and ML for sure....they are the future

I mostly agree. CGPA is not everything, but pretending it doesn't matter is bad advice. For campus placements, CGPA is often the first filter. A student with an 8+ CGPA simply gets access to more opportunities. That said, CGPA gets you the interview. Skills get you the offer. For average students: improve both. For exceptional students: strong skills can sometimes compensate for a lower CGPA. For everyone else: don't give companies a reason to reject your application before they even see your resume. 🎯

When I was in my B.Tech, every senior used to say that CGPA is not that important. But the ones who said this were usually the ones who didn't study. When the time for placements comes, CGPA does matter. In fact, many companies use CGPA as a filtering criterion, and for some companies, the requirement is a CGPA of 8.0 or above. However, I can say that for those who are exceptionally good at coding and other technical skills, CGPA may not matter much. But if you are an average student, CGPA does matter. Having a good CGPA increases your chances of getting a job during campus placements, as you become eligible for more opportunities. What do you think about this ?

@Dhruvam987 Interesting. What makes you choose AI/ML over something like sales or distribution?

@devfrom_hyd Maybe ai and ml

@Its_Nova1012 Codex

Be honest, which AI coding tool will you choose to build your SaaS?

@Dhruvam987 Then: “overnight success” according to everyone who missed the first 5 years. 😄

Employee: 3k → 3k → 3k → 3k → 3k Freelancer: 1k → 0.5k → 2k → 6k → 1k → 8k Entrepreneur: 0 → -1k → 0 → 5k → 20k → 100k → 500k Startup Founder: -5k → -20k → -80k → investor rejections → sleepless nights → people calling you crazy → almost giving up... Then suddenly: 1M+ Most people want the last number. Very few are willing to survive the numbers before it. Which path are you choosing? A) Employee B) Freelancer C) Entrepreneur D) Startup Founder

@jingconan If no: retry it, parallelize it, cache it, scale it. Read and write operations shouldn't be treated the same. ⚡

Most agent frameworks are missing a critical distinction. They treat all “skills” the same. That’s a mistake. In real systems, one thing matters a lot: → does this change state? If yes: - retries can break things - ordering matters - failures leave partial state If no: - you can retry freely - run in parallel - keep things simple This is why most agent demos work — and most real systems break. I wrote a simple model for thinking about this: skills = (what it describes) × (how it executes) And tried to make it concrete here: → github.com/DeepVista-AI/d… Full write-up: → go.deepvista.ai/agent-skills-2d