IIITDM Jabalpur Placed Under 25%. The Offshore Pyramid Just Broke.

If you have been running an India fresher pipeline for the last decade, the math you used to plan headcount no longer works. NASSCOM and EY data published in April and May 2026 confirm what campus coordinators have been whispering since Diwali: the entry-level share of tech hiring in India fell to 15% in 2025 from 28% in 2024, and at brand-name institutes like IIITDM Jabalpur, fewer than 25% of the May 2026 graduating class had an offer in hand at the gown ceremony.

This is not a soft year. This is a structural break in the offshore graduate pyramid, and it changes where you should be looking, what you should be filtering on, and how much you should trust campus tier as a signal.

The two "15%" numbers are the whole story

The number sourcers are quoting in Slack is 15%. The problem is that there are two of them, and the overlap is what actually matters.

The first 15% is demand-side. NASSCOM and EY's joint read on FY26 puts entry-level roles at 15% of all new tech hires, down from 28% the prior year. EY's separate analysis estimates entry-level IT roles have already declined 20 to 25% from automation alone. NASSCOM president Rajesh Nambiar has framed it directly: the sector is moving "from volume hiring to deliberate talent pool creation aligned with future capabilities."

The second 15% is supply-side. Kapil at Quess estimates roughly 1.5 million engineering students graduate in India every year, and only about 15% understand emerging technology and are job-ready, concentrated in Tier-1 colleges with updated curricula. Xpheno's Kamal is blunter: mass hiring at entry level in IT will no longer be a reality.

Stack them and the picture clarifies. The market is now buying roughly 15% of grads, and roughly 15% are job-ready. Those are not the same 15%, but they overlap heavily, and that overlap is essentially the entire viable fresher market in India today.

2.3%

What broke the pyramid

For twenty years, the offshore math was simple. TCS, Infosys, and Wipro absorbed tens of thousands of freshers a year, rotated them through training programs in Mysuru or Pune, and billed them out at a margin. That funnel fed everything downstream, from GCC poaching to product company hiring two years later.

That funnel is closing. TCS cut more than 12,000 jobs last year citing market unpredictability and AI disruption. Infosys and Wipro followed with hiring freezes and reduced intake. Estimates indicate AI can already perform 20 to 40% of common tech functions including coding, generating test cases, and maintaining documentation, which is exactly the work the rotation programs were designed to staff.

And the displaced freshers are not competing only with each other. Layoffs.fyi recorded over 73,000 tech worker layoffs across 95 companies in just the first four months of 2026, versus roughly 125,000 in all of 2025. Juniors looking for their first role are pitted against laid-off mid-levels with two years of production experience and the same salary expectation.

The funnel did not slow. It closed, and the people inside it scattered into four destinations your ATS does not index. </pull>

pull The funnel did not slow. It closed, and the people inside it scattered into four destinations your ATS does not index.

## Campus tier is now a lagging indicator

If you still rank-order by campus tier, you are sourcing from a degraded signal. IIITDM Jabalpur is a brand-name institute. A 400-student class with fewer than 25% placed is not a tier problem, it is a market problem, and many of the 75% un-placed at Tier-1 schools are stronger than placed students at Tier-2 and Tier-3 schools where local recruiters never stopped showing up.

The conventional pyramid is inverted at the margin. The student who got placed at a third-tier campus into a body-shop role is now in the "placed" column on every spreadsheet your TA team builds. The IIITDM Jabalpur student with three Kaggle medals and a merged PR into a Hugging Face repo is in the "unplaced" column. If you are filtering by placement status, you have it exactly backwards.

This is why plain-English sourcing matters more than it used to. Boolean over Naukri or LinkedIn does not give you "IIITDM 2026 grads with merged PRs in open-source ML repos," which is why we built [Refolk](/): you describe the person in plain English, and you get a ranked shortlist across GitHub, LinkedIn, and the open web, including the un-placed candidates the placement cell never logged.

## Where the displaced cohort actually went

The 75% at IIITDM Jabalpur did not disappear. They scattered, and the four destinations are roughly predictable.

### 1. Higher studies and M.Tech

The default fallback. CAT and GATE registrations are up, and admissions counselors at IITs and IIITs report a flood of applicants who would have taken a TCS offer two years ago. These candidates will resurface in 2027 and 2028 with a master's and a much higher salary floor. Worth tracking if you have a two-year horizon.

### 2. Extended internships at product companies

GCCs in Bengaluru and Hyderabad have quietly extended internships into six and nine-month contract roles. Google, CRED, Palo Alto Networks, and UKG all visited IIITDM Jabalpur's 2025 placement drive. The students who interned but did not convert to FTE are still in the building, on contracts, doing real work, and not showing up as "employed" on any campus dashboard.

### 3. GCCs that bypass Indian IT services entirely

This is the biggest shift. Foreign multinationals are hiring directly into their India captives, skipping the TCS-Infosys rotation entirely. The candidate goes from campus to a Google or Walmart Labs badge with no services firm in between. Your old sourcing playbook, which assumed two years at Infosys as a filter, now filters out the strongest candidates.

### 4. AI-adjacent specialist tracks

MLOps, data engineering, security, and applied ML. A joint NASSCOM-Deloitte analysis estimates India's AI talent pool could reach 1.25 million by 2027, with demand likely outpacing supply. The students who saw the writing on the wall in 2024 reskilled into these tracks and are now the most over-subscribed segment of the market.

The BPM counter-signal nobody is pricing in

One number cuts against the gloom. Naukri's January 2026 JobSpeak Index showed fresher hiring in the BPM segment rising over 30% year-on-year, even as overall hiring stayed selective.

That is not a recovery. That is repackaging. Routine work is being moved to lower wage bands rather than eliminated, and BPM is absorbing the volume that used to flow through IT services rotation programs. For recruiters who need junior headcount at scale, BPM-to-engineering crossover candidates (people who took a BPM role in 2025 to pay rent and kept shipping side projects) are a real and underpriced pool.

You will not find them on a campus visit. You will find them by searching for active GitHub accounts with a BPM employer on LinkedIn, which is the kind of cross-source query Refolk handles natively and that Boolean cannot express.

The "AI washing" caveat

Before you over-correct, hold one fact in view. Anthropic's Dario Amodei has predicted the end of software engineering by 2027. Marc Andreessen and others argue AI is a fig leaf for correcting pandemic over-hiring, a practice now called "AI washing." Both can be partly true.

If the cuts are 60% structural automation and 40% AI washing, the offshore pyramid does not come back, but the floor stops dropping in late 2026. If it is the reverse, fresher hiring partially recovers by FY28 and the firms that built deliberate sourcing capability through the downturn will have a two-year head start on the firms that simply paused campus visits.

Either way, the move is the same: stop treating campus placement status as a quality signal, and start sourcing on verifiable work product.

67%

What this means for your sourcing playbook

A few concrete shifts for engineering leaders and TA teams running India pipelines into Q3 and Q4 2026.

Stop ranking by campus tier alone. Use it as one signal among several. A 2026 IIITDM Jabalpur grad without a placement is statistically more likely to be strong than a placed Tier-3 grad, and your filters should reflect that.

Index on work product, not employer history. GitHub commit history, Kaggle rank, Hugging Face contributions, and personal projects are the only signals that did not degrade in 2025. Refolk's index already cross-references these against LinkedIn and the open web, which is roughly 98,500 India-based engineers with Python or JavaScript on file, concentrated in Bengaluru, Noida, Mumbai, Kochi, Ahmedabad, Pune, and Delhi NCR.

Map the GCC alumni graph. The strongest 2024 and 2025 grads are sitting in GCC contract roles right now. They are not on Naukri. They are reachable through GitHub, conference talks, and second-degree LinkedIn connections from current GCC employees.

Build a BPM-to-engineering pipeline if you need volume. The 30% YoY BPM fresher growth is hiding a pool of candidates who took a paycheck job and kept building. Filter for active GitHub or open-source activity alongside a BPM employer.

Stop scheduling campus visits as the primary motion. Visits still matter for employer brand. They are no longer the primary sourcing channel. Plain-English search across GitHub, LinkedIn, and the open web is.

The offshore pyramid is not coming back in its 2019 form. The 15% who are job-ready and the 15% the market is buying are now essentially the same group, and the firms that learn to find them outside the placement cell will run circles on cost and quality through the next two years.

FAQ

Is the 15% entry-level hiring figure tech-only or all industries?

The 15% figure comes from NASSCOM and EY's read on India's tech sector specifically, comparing 2025 to 28% in 2024. The India Skills Report 2025 found a broader cross-industry intent-to-hire-freshers figure of 14% in 2024, down from 18.8% the prior year. Tech is leading the contraction, but it is not isolated, and the BPM uptick noted in Naukri's January 2026 JobSpeak Index is the main counter-signal worth tracking.

How do I source the un-placed Tier-1 cohort if they are not on Naukri?

They are on GitHub, on Kaggle, in Discord servers for open-source ML projects, and on personal portfolio sites. They are also on LinkedIn but without a "current employer" field, which makes Boolean queries painful. A plain-English query like "2026 IIIT grads in India with public ML repos and no current full-time role" works better, and that is exactly the kind of cross-source sourcing Refolk is built for.

Will fresher hiring in India recover in FY27 or FY28?

Probably partially, but not to 2024 levels. NASSCOM-Deloitte's 1.25 million AI talent pool projection by 2027 implies demand for specialist freshers in MLOps, data, and security will keep growing. Generalist services-firm rotation programs are unlikely to return at scale. If you plan headcount assuming 2024 ratios will come back, you will over-hire generalists and under-hire specialists.

Are GCCs really hiring freshers directly now?

Yes, and increasingly so. GCCs in Bengaluru and Hyderabad have shifted from poaching two-year services-firm alumni to direct campus and contract-to-hire, partly because the services firms are not training at volume anymore. Google, CRED, Palo Alto Networks, and UKG were among the firms named in IIITDM Jabalpur's 2025 placement drive, and the pattern is consistent across the top dozen IIITs and IIITDMs.