Explore Across Campuses

the L1 support engineer. the KYC analyst. the manual QA tester running test cases. the compliance reviewer. those are the jobs 300,000 Indian graduates used to enter tech through. AI is eating that layer first and nobody is building a bridge over it. what is the first rung of the ladder now

Oracle cut 21,000 jobs globally this year, around 12,000 of them in India. then in September another wave, 3,000 more here. what gets me is the method: system access off at 4am, Slack disconnected by 5:30, email at 6. you find out by failing to log in to your own laptop

dell's zinnov research says 55% of routine work inside indian GCCs is exposed to AI automation, and nearly 60% of the workforce will need reskilling by 2030. that is not a skills problem. that is a demand problem. the work those people were doing is being done by something that does not need reskilling

90% of devs still review or validate AI output. 79% require human approval before production. the cost is the top barrier to broader agent use at 30%, ahead of integration at 20% and governance at 20%. we are not short of oversight. we are short of budget

an AI/ML hiring jump of 260% sounds like a boom. the same data says 49% of that demand is for 3-5 years experience. the boom has an age requirement. if you are 22 the growth rate is not a number about you, it's a number about the market you will enter in four years

everyone posting 'AI took my job' from 2023 with a fake LinkedIn banner is why this conversation has no credibility. AI didn't take anyone's job in 2023. AI took nobody's job. it was a content cycle. this year it took 175,000 and that is not content, that is a quarter

students and colleges are both still working off the 2015 model: do a degree, get a package, four years. the contract is broken at both ends. companies stopped buying the output of that contract and colleges kept selling it. both were behaving reasonably. that is what makes it hard

cool so i spent 2 years learning DSA to get a job and now the job wants AI and the AI needs 6 months of experience that nobody gave me. cool. love this for us. the pipeline has a gap in it and the gap is where my entire cohort is standing

55% of GCC routine work exposed to AI. bro that number should be printed on a banner at every campus and NOBODY is printing it because it would scare the placement season. we are being told a 4-year plan in 4 years of memes

a lot of us are in the same place: not unemployed, not safe. watching a number every quarter that is no longer assumed to go up. that is a specific kind of fear that has no name and no advice column. i have not found anyone who is handling it well. that might be the most honest thing i can post today

90% of the code we ship has no human touch. that's what a founder of a company building AI agents said on stage in mumbai this month. 30,000 businesses use their agent product daily and india is 8-10% of revenue. read that number again: the largest talent pool on earth is 10% of the market for the thing that is replacing entry-level coding

every few years this industry runs a cycle where a tool is going to end the industry. spreadsheets. the internet. low code. agents. the tool does not end the industry. it ends a job description. the industry then takes four years to write the next one and complains the whole time that nobody is ready. that is the pattern and we are in year one of the current one

university placed 340 students. 300 of them to mass recruiters at 3.6 LPA. 40 to companies with real names. 12 to roles you'd put on a LinkedIn headline. that's the pyramid and only 3% of us are ever told which part we're in until the last week

if AI removes entry-level execution work, the next effect is that organisations also remove the rung people used to climb. not just the job. the path. you cannot promote from a floor that no longer has a first step, however many training programmes you buy

the boundary between writing code, testing it and deploying it is dissolving. everyone knows this. what nobody has said clearly: the person who used to be accountable for that whole chain is now accountable for a chain where two of the three links are autonomous. that is a different job and most job descriptions still describe the old one

for the last three years the argument has been whether placements are declining. the answer from the data is yes, at every tier, in the public numbers. what hasn't changed is the assumption in the classroom. we know. we keep teaching to the old number because the old number is what the parents heard

asked my manager in august whether my role was at risk. he said 'no, we have a plan for everyone.' in september the same team lost two people. in october a third. i am still here. the plan for everyone turned out to mean everyone who was already here in august

there is a version of this industry where 2026 is a great time to be in tech and there is a version where it is a bad one, and they are happening to different people in the same city on the same day. both are true. any post that only tells you one of those is a post with an agenda

severance capped at 26 weeks of base. that's the number that was in the packet. i know because i read the packet. 26 weeks for a person who gave 9 years is a decision someone made in a spreadsheet

the cuts hit people with strong performance ratings. that's the detail nobody is writing about. if your review was excellent, that is not a shield, it is evidence. being good at the job you were being removed from is now on the record