Bulk Hiring in Indian IT: The Bottleneck Moved

A mid-sized Indian IT services company hiring 100 freshers receives 5,000 to 10,000 applications. That ratio has not changed much since 2022. What has changed is everything around it — TCS narrowing fresher offers from 44,000 to 25,000, Infosys tightening intake to specialist tracks, and a sudden GenAI-skills gap layered on top of the same flood of resumes (alp.consulting, Top 8 Tech Hiring Challenges in India 2026). Bulk hiring in 2026 is not broken because of sourcing volume. It is broken because the screening layer that sits between sourcing and offer was never designed for this.

We have spent the last several months running pilot screening campaigns at the 2,500–3,000-candidate scale with Indian enterprises like HyperVerge. The pattern in every conversation with TA leaders is the same: the resume pile arrives, the HR team has eight working days to evaluate it, and the math does not close.

What is actually different in 2026

The headline change is selectivity, not volume. Bulk hiring is gradually narrowing in many campus recruitment programmes as companies move toward skill alignment and job-ready evaluation rather than mass intake. India now reports a roughly 90% shortage in GenAI-ready talent and a 55–60% deficit in cloud-native expertise, per the same alp.consulting analysis cited above.

Three concrete shifts show up in the funnel.

The first is that fresher intake has narrowed but applications-per-seat have climbed. When TCS cut its fresher pool by roughly 40%, the same campuses kept producing graduates — every remaining seat now attracts more applicants, not fewer. The funnel got steeper, not wider.

The second is that AI-generated resumes have made the resume stack less useful as a signal. Polished resumes look near-identical at the top of the pile. Recruiters cannot tell from a PDF who can hold a real conversation about the work they would actually do.

The third is that the niche-skills shift means HR is not just sorting for “any decent fresher.” They are sorting for cloud-native, GenAI-aware, AI-ops-ready specialists — and the assessment signal for that does not fit on a resume. It needs a conversation.

Where FY26 fresher plans break

When LTIMindtree announced plans to hire 5,000 freshers in FY26 (HR Katha, 2025), the conversation in their TA team wasn’t about where to find candidates. India’s engineering college system produces candidates at a scale the market can’t absorb. The conversation was about how to run first-round screens across 40+ campuses with a TA team sized for a fraction of that volume.

That mismatch — between the volume of hiring intent and the throughput capacity of HR — is the defining constraint in bulk IT hiring in 2026. It’s a logistics problem, not a talent problem.

Take a mid-sized IT services firm targeting 3,000 fresher hires this season. The screening funnel for fresher roles typically runs at a steep ratio — many candidates for every eventual offer. That means tens of thousands of first-round interactions, each requiring scheduling coordination, recruiter bandwidth, and candidate follow-up.

At any reasonable estimate of time-per-screen, the queue runs into weeks, not days. Well past the window when placement seasons close and candidates accept competing offers. Nobody runs this math before the season starts. It surfaces during week three, when recruiter no-show rates are climbing and the shortlist is still 40% of target.

The standard fixes don’t hold at this scale. Contract recruiters need calibration time before they’re reliable. Extended timelines cause candidates to accept faster-moving employers. WhatsApp nudge campaigns add coordination noise without solving throughput. These are tactical responses to what is structurally an engineering problem.

Why IT services is harder than it looks from the outside

Consumer tech companies typically run one consolidated hiring drive with a few hundred slots across a handful of tier-1 colleges. IT services firms operate differently. They simultaneously source from IITs and NITs for high-complexity delivery roles, deemed universities for mid-tier volume, and state engineering colleges for the bulk of their headcount. Each campus runs on a different calendar, has a different placement officer with different expectations, and produces a candidate pool with meaningfully different preparation levels.

This multiplies coordination overhead multiplicatively, not linearly. A recruiter managing phone screens across three concurrent campuses is context-switching every 45 minutes while handling rescheduling, placement officer check-ins, and candidate queries in parallel. That’s not a sustainable operating mode at 3,000-candidate scale.

Infosys and TCS solved this historically through institutional infrastructure — a dedicated campus function built over years, pre-built assessment frameworks, deep college relationships. For IT services firms outside the top five by revenue, that infrastructure doesn’t exist. Bulk hiring for them is always somewhat improvised, and that improvisation breaks first at the screening stage.

The bottleneck moved. Most ATS stacks did not.

Here is the uncomfortable part for vendor evaluations this year: every ATS in your evaluation deck was designed when “bulk hiring” meant “track this many candidates through a pipeline.” The screening conversation was either outsourced to recruiting agencies or handled by HR over the phone, one candidate at a time.

When the screening cost is ₹85–150 per 10–15-minute phone screen and you have 5,000 applicants, the unit economics break before you start. A traditional ATS can show you the 5,000 applications. It cannot tell you which 500 are worth a real conversation. That is a different product, and your stack does not currently include it.

On one pilot drive of roughly 3,000 applicants, HR screening time fell from about 18 hours of manual profile review to 1-2 hours of dashboard review. The technology shift was not sourcing. It was that the screening conversation moved from a human phone screen to a 15–20-minute AI voice agent running 20 parallel slots per 15-minute window.

Candidates self-scheduled via a link sent after application, called in at their chosen time, and completed the conversation with no recruiter involvement in real time. HR reviewed the structured output on a dashboard and rejected with confidence. An 89% reduction in HR screening time, compared to manual profile review, was the measured outcome.

That is not faster phone screening. It is a different category of operation, and it removes the scheduling-and-screen step entirely from the recruiter’s plate.

What changes when the screening layer changes

Three operational outcomes you can plan for if you replace the phone-screen layer with a structured AI conversation.

You stop rejecting good profiles. The biggest cost in high-volume hiring is not the time spent on bad candidates — it is the false rejections you never measure. When HR has 30 seconds per resume, the candidate with an imperfect resume but strong communication gets cut by accident. A voice screen surfaces that signal directly, before the rejection is final.

You compress the cycle from weeks to days. In our pilot, the screening phase went from 6-10 working days to 1-2 working days. That matters when campuses have rolling placement windows and the offer that lands first wins.

You free HR for the decisions that need humans. Eight days of phone-screening burns out the people you most need fresh for the offer and joining-decision conversations. AI screening does not replace HR judgement — it gives HR their judgement time back.

Borderline candidates still get human review, and final offers involve direct conversation. But for the clear no-gos — failed mandatory qualifications, communication significantly below role threshold, mismatched location expectations — the system can surface decisions without recruiter time investment. For a deeper breakdown of the capacity math, see hiring 5,000 freshers as a capacity problem, not a sourcing problem.

The honest constraint: build per-JD screeners first

None of this works unless per-JD screener rubrics are built before the drive launches. A Java developer screen, a business analyst screen, and a technical support screen have different first-filter criteria, different communication benchmarks, and different weight on problem articulation versus domain knowledge. A single generic screener deployed across roles doesn’t outperform a generic phone screen — it just fails at higher volume.

The IT services firms getting real efficiency in 2026 are the ones spending time per role-cluster building screener frameworks before campus season starts. That upfront investment is what makes the throughput math hold.

The compliance overhead that arrived without warning

Several large global clients now require their IT services partners to demonstrate bias-auditable hiring practices for offshore delivery roles. The EU AI Act’s high-risk classification for AI used in hiring decisions (EU AI Act Article 6) is starting to appear in vendor contract language, even for Indian entities with EU-headquartered clients.

This adds a documentation layer to bulk hiring that didn’t exist five years ago. Every AI-assisted screening decision needs an audit trail: which rubric, what score, what threshold triggered the outcome. Most TA teams discover this requirement after deployment rather than before it. Building the audit trail architecture during screener setup is significantly cheaper than retrofitting it when a client asks for documentation post-drive.

The actual planning failure

Most IT services TA leaders budget for sourcing — job boards, campus connect fees, college visit logistics — and treat screening as a zero-cost step absorbed by the existing team. The result: the top of the funnel fills, and candidates stall mid-pipeline for weeks while recruiters work through the queue one phone screen at a time.

For a clear look at how the recruiter-to-candidate ratio math compounds across teams at different sizes, the 1:30 HR-to-candidate ratio breakdown maps the problem precisely.

India’s top IT services firms alone hired more than 80,000 freshers in FY25 (NASSCOM). The firms that close faster in 2026 are the ones that treat screening as a logistics operation — staffed, tooled, and load-tested before the drives start — rather than an ad hoc activity that runs off recruiter goodwill. The capacity ceiling is real. Plan around it before the season starts, not during it.

The 2026 question worth asking

The right question for an Indian IT services TA leader this year is not “which ATS should we buy?” It is “what does our screening layer look like when applications-per-seat keeps climbing and our HR headcount does not?”

If the answer is “phone screens by HR,” our read when we first wrote this was that the math would break by Q3. If the answer is “an AI conversation that plugs into our existing ATS and produces a structured rejection-or-shortlist signal,” the math closes — at a cost well under the ₹85–150 of a phone screen, with the second-order benefit that the team is not burnt out by Diwali.

We built HireQwik because every TA director we spoke with was solving the same equation by hand. If you would like to see what the unit economics look like for your campus drive size, book a demo and we will run the numbers with you.