The Fresher Pyramid Is Being Dismantled

Sridhar Vembu told BW Businessworld that Zoho has not hired many software engineers in two years and will not restart until its AI and vertical strategies are clear. He wants engineers with deep domain expertise in healthcare, automotive or insurance, not pure coders, and says even existing engineers must spend more time with customers. The comments land against hard data: Naukri JobSpeak shows IT hiring down 3% YoY in June while AI roles rose 16%, TeamLease EdTech pegs fresher IT hiring intent at 76% for H2 2026, and Quess reports GCC talent gaps of 43% in AI and data analytics. Zoho is also pushing into chip design, robotics and medical equipment.

Teksands take: The loudest hiring signal this fortnight is a company that stopped hiring. Zoho, profitable and unlisted, paused engineer recruitment for two years to rebuild around AI and hardware, and Vembu is openly telling young engineers that pure coding is no longer the safe path. Meanwhile Naukri shows IT hiring down 3% YoY with AI roles up 16%, and GCCs report 43% talent gaps in AI and data analytics. The market is not short of engineers. It is short of engineers who understand an industry. TA teams still screening on frameworks and years of experience are fishing in the pond that is draining.

An ANSR and T500 Research report, reported by ET CIO on 24 September 2026, says India's GCCs are moving to a 'Workforce 2.0' model: engineering pods of three to five professionals supported by 50-100 AI agents, replacing the traditional broad-based junior pyramid with a diamond-shaped structure weighted toward mid-level AI-enabled engineers. Analysis of 5,000-plus live openings and 20,000-plus roles hired in the past year shows 62% of hiring concentrated in the 3-8 year band, median experience of six years, and nearly 65% of new positions requiring AI skills. GCC hiring grew 12-15% year-on-year in H1 2026, faster than the broader IT/ITeS sector, with AI role demand up about 45%.

Our take: The pyramid is being replaced by a diamond, and most TA teams have not noticed. If 62% of GCC hiring sits in the 3-8 year band and 65% of new roles demand AI skills, then fresher-heavy pipelines and role-based requisitions are the wrong infrastructure. Rebuild around capability pods, mid-career AI sourcing and retention, and treat flex talent and AI agents as part of the workforce plan, not a procurement afterthought.

22 companies just lost IIT campus access for two years

The All IITs Placement Committee, meeting at IIT-Guwahati this month, decided to stop cooperating with 22 companies for the next two placement seasons after they failed to honour offers made to 2026 graduates. The number of affected offers exceeds 150 across 23 IITs, with Oracle alone linked to around 40 withdrawn offers. Companies were given a September 15 onboarding deadline; letters go out by September 30. The AIPC is deliberately avoiding the word 'blacklisting', calling it 'non-cooperation' instead. Last year only 10-11 companies reneged. Placement offices have circulated affected students' CVs to existing recruiters, and roughly 20-30% of those students found roles through cross-referrals.

This is the entry-level talent market's version of a credit rating. Companies that rescind offers are now visibly downgraded, and the penalty is access to India's strongest fresher pipeline for two full placement cycles. For TA leaders, the practical implication is blunt: your offer-to-joining ratio assumptions need a buffer, your headcount sign-off needs to happen before you go to campus, and your campus brand is now a balance sheet item. The firms that honour every offer will quietly win the next three seasons.

Four signals worth watching

The lesson: This is a hiring-controls story dressed up as geopolitics. The contractor cleared onboarding with fake documents and a proxy holding the laptop. Indian GCCs and IT services firms run thousands of remote contractors on similar paperwork. If client security teams read this, expect contractor identity verification and in-person checks to move from 'nice to have' to audit condition. Fix it before a client asks.

India is exporting skills while the US entry-level tech ladder breaks

Hiring for AI-native and domain-plus-AI roles?

Teksands works with GCCs, IT services firms and product companies to source mid-career AI, data, platform and domain-plus-technology talent — the 3-8 year band where the market is tightest. If your requisitions are shifting from volume fresher intake to specialist capability, we can help you map the talent, benchmark the compensation and shorten the time to close.

Split your requisitions into three tracks this week

Take your open roles and sort them into three buckets: AI-native or domain-plus-AI roles, mid-career 3-8 year roles, and generalist roles. Set separate compensation bands, sourcing channels and assessment criteria for each, because the evidence says these are now three different markets. Then check whether your fresher intake plan still assumes the entry-level demand curve that Zoho, the GCC pod model and the US graduate data all suggest is flattening. If it does, rewrite it before the next campus cycle locks in.