I spoke to a seventh-semester computer science student in Mangalore recently. I will call her Amrita, which is not her real name.
She had tried job portals, referrals, cold emails, and LinkedIn. She had a three-month window for an internship.
Some of the listings she found were unpaid. Others barely explained the work. Her campus was pushing her towards roles that didn’t always use the technical degree she’d spent years working towards, including sales and marketing.
Her question was simple: what was the point of learning all this technical material if the first available job did not use it?
I’ve had hundreds of conversations over the past two months. Amrita’s account is one of them. It can’t tell us how the whole internship market works.
What bothered me was how much she had to figure out alongside the search. She was expected to find a real opportunity and judge it, with little guidance on how hiring or professional relationships work.
Hiring figures don’t tell a student where to apply
The hiring reports I read don’t describe an empty market.
Business Standard’s September report on Naukri’s August 2026 JobSpeak data recorded a 14% year-on-year rise in white-collar hiring activity. Hiring activity for candidates with up to three years of experience rose 15%. That is useful evidence of demand, with an important limit: JobSpeak measures job postings and recruiter searches on Naukri. It does not measure actual hires, gig work, or campus placements.
The picture changes depending on where a student studies and what they can show. Deloitte India’s 2026 campus workforce survey, covering 257 organisations and 505 campuses, reported larger campus hiring budgets and stronger internship-to-offer conversion. It also reported a sharp institute-tier gap. Its median B.Tech compensation figure was ₹18.7 lakh at the top 10 institutes and ₹5.75 lakh at tier 3 institutes.
Those figures leave room for an exhausting search. A student without a brand-name campus, referrals, or work they can show may struggle to reach the employers behind the numbers.
Recent reporting points to the same tension. The Economic Times reported that Aon’s 2026-27 campus study found 51% of surveyed organisations expected to hire more interns and 36% expected to hire more engineers. The report also described employers as more selective about specialised skills and job readiness. That is a hiring signal, not a guarantee that a particular student can get a useful role before graduation.
I can’t turn those reports into a useful listing for Amrita. She still needs to know which company to trust and who might read her message.
What the AI findings can tell us
Amrita and her friends see AI everywhere. They see companies asking for AI skills. They also see fewer obvious junior openings and wonder if tools have removed the need for freshers.
Deloitte found that 35% of surveyed organisations said AI adoption was creating new entry-level roles or skill requirements. Employers also use AI in screening and assessment. A student may encounter it while applying, before they get a chance to use it at work.
It does not prove that AI caused every slow hiring process, cancelled internship, or bad entry-level offer. Companies hire and pause hiring for many reasons: revenue, client demand, team plans, budgets, and the time it takes to train someone. A student who believes that every door is closed by AI may stop trying doors that are still open.
Then there’s access. Watching a video about an AI coding tool gives you little practice with its mistakes. You need time using it to notice when context runs out or an answer needs checking. A monthly subscription can be a serious cost while you’re still in college.
Use whatever practical access you can get. That might be a college lab account, an open-source model, a free tier, hackathon credits, or a small team with a real problem to work on.
Then write down what happened. For example, if you traced a failing API request with a tool, explain how you checked its answer against the logs. Include what you fixed. Someone reading can then judge the work instead of guessing what “I used AI” means.
An internship needs a clearer test
I understood Amrita’s caution about new companies with unclear scopes. An unfamiliar name alone doesn’t establish fraud. New companies can offer excellent work, but you need to find out more before accepting.
I would ask for these six things in writing:
- What problem will I work on in the first two weeks?
- Who is my manager, and how often will we meet?
- Is there a stipend, and when is it paid?
- What tools, accounts, or equipment will I need?
- What does a good final output look like?
- Does this role promise a full-time offer, or only give experience?
The sixth question matters because an internship is not automatically a job pipeline. In its answer on the Prime Minister’s Internship Scheme, the Ministry of Corporate Affairs said the scheme aims to provide training, experience, and skills. It also said an internship offer is not a promise of employment during or after the placement.
Students also need a basic safety rule: do not pay to apply for a job or internship. The AICTE National Internship Portal says that registration and applications on its portal cost students ₹0, and that a person asking for payment to apply is a scam warning. Verify a company through its official domain, the people who work there, and a direct call or email. A polished PDF is not verification.
The missing subject is professional practice
Many students learn data structures, operating systems, and databases. They may never be taught how to write a useful cold email, turn a small project into a public case study, ask an alumnus for 15 minutes, or check whether an internship has a real manager.
That leaves them at a disadvantage against students who learned those things at home, through seniors, or through an earlier internship.
After our call, I asked Amrita to send me a short note with the decisions we’d made. That would give us something specific to follow up on.
The same idea works for a job search. After every application, keep a small record:
Company:
Role:
Person I contacted:
What I can show them:
Next action and date:With that record, you can see which applications need a follow-up and what you’ve already tried. You also have something to write about. A short series on the search could explain what you checked and which questions remain unanswered. You don’t have to present yourself as an exceptional programmer to make that useful.
If posting under your name feels too exposed, you could start a separate professional account for practice. Keep it honest. Read people whose work you respect, try their format with your own evidence, and credit the source of an idea you borrow.
An event or technical group can give you someone to ask. A careful project note gives them something to respond to. Neither guarantees a role, but both are actions within reach.
I’d like colleges to practise this with students before internship season. Review a real listing together. Help draft a message to an alumnus. Ask who will follow up after a call. I wrote about that last question in mentoring that ends with a next step.