The Hyderabad BSc graduate and the resume stack

Arjun completed a BSc in Computer Science from a private college in Kukatpally in 2024. The college is not a bad one. It is not JNTU. It is the kind of college that appears in Hyderabad's institutional geography as one of roughly four hundred affiliated colleges where a BSc-CS degree costs around ₹60,000 a year in fees, produces students with functional Python knowledge and a reasonable understanding of databases, and has a placement cell that makes cheerful promises in the prospectus and delivers, in practice, two or three walk-in recruitment drives per year from companies whose names you have not heard and whose packages start at ₹2.4 lakh per annum. This is not dishonesty, exactly. It is the gap between what a private affiliated college in Kukatpally can credibly promise and what the IT recruitment economy in Cyberabad actually wants.

Arjun had been applying for jobs since March 2024, when he finished his final semester. By January 2026 — ten months into his job search — he had sent out four hundred and twelve applications. He had tracked them in a notebook, because his phone's storage was insufficient for a spreadsheet app and because he is, by nature, a systematic person. The breakdown, as he described it to me, was roughly: three hundred and twenty on Naukri, fifty-one on the Apna app, twenty-two on LinkedIn (where he had a profile but no paid subscription), and nineteen on various company career pages he had found through Google. He had received sixty-seven acknowledgement emails, eighteen shortlist calls, twelve telephonic screening rounds, seven in-person or video interviews, and zero offers. He had not, at any point, received a reason for any rejection.

This is a story about those four hundred and twelve applications, about the specific economy of the Hyderabad IT recruitment corridor, and about what changed in month eleven when his college friend Deepak — who had himself gotten a job six months earlier and felt a complicated guilt about it — installed a small Telugu-speaking AI on Arjun's phone.

🗓️ What the Naukri economy looks like from the inside

There are, depending on who is counting and what the threshold is, between twelve hundred and two thousand IT and IT-adjacent companies operating in Hyderabad's tech corridor — the stretch from Hitech City through Madhapur to Kondapur to Gachibowli, with satellite clusters in Kukatpally and Uppal. The big ones — the Infosyses and TCSes and Wipros and HCL Technologies, the Amazons and Googles and Microsofts, the mid-tier NASSCOM members — hire through structured campus processes, referral pipelines, and internal transfers. The walk-in-job-board market that Arjun was navigating is a different layer entirely: small service companies, BPO support arms, IT staffing firms, and the assorted KPO/data-entry-adjacent roles that account for a meaningful fraction of Hyderabad's tech employment but are entirely distinct from the Cyberabad skyline that appears in the newspaper photographs.

In this layer, the economics work like this: a recruiter receives three hundred resumes for a ₹2.8-lakh-per-annum junior developer role. She has thirty minutes. She scans the first twenty that the Naukri algorithm surfaces, shortlists six for a call, interviews three on the phone, and sends two to the hiring manager. Of the original three hundred applicants, two hundred and ninety-eight receive nothing. Not a rejection, not a status update — just silence. The Naukri algorithm surfaces resumes based on a combination of keyword matching, profile completeness score, and how recently the resume was updated. A resume last updated four months ago, with standard college-level keywords, from a college that the algorithm doesn't have in its preferred-institutions list, will be ranked below a resume last touched yesterday from someone who has added the word "Agile" in a project description.

Arjun's resume had not been updated since August 2024. It was two pages long. It listed his BSc, his college, a course in Python from NPTEL, a project on a library management system that he had built in his third year, a project on a basic inventory tracker he had done as a freelance exercise for a textile shop in Nizamabad, and five bullet points of skills including Python, SQL, HTML, CSS, and "basic knowledge of React." It was an honest resume. It was the kind of resume that a good student from a mid-tier Kukatpally college, applying in good faith, would naturally produce. It was also a resume that the Naukri algorithm was, for structurally comprehensible reasons, ranking below six thousand other resumes every week.

He did not know this. Nobody had told him. His college's placement cell had reviewed his resume once, in October 2024, and said it looked "fine." His father, a police sub-inspector, had no frame of reference for IT recruitment. His mother was occupied with his grandmother. Deepak, his college friend who had gotten a job, had gotten it through a reference from a relative at an IT services company — a pathway that is, in Hyderabad's tech recruitment ecosystem, the single most reliable one, and also the one that cannot be manufactured on demand.

1. 📋

   March 2024 — First Naukri application

   BSc CS from a Kukatpally college. Resume uploaded August 2024. Naukri profile score: 61%. Keywords missing: Django, REST API, Git, Agile. Ranked below 6,000 resumes.

2. 🛑

   January 2026 — 412 applications, 7 interviews, 0 offers

   67 acknowledgements, 12 screening calls, 7 interviews. Spoken-English section decisive in every failure. Zero feedback from any rejection.

3. 📱

   January 2026 — Deepak, from Kondapur

   College friend, employed via referral, sets up a Telugu-language AI on Arjun's phone. Agent reads resume: 61% profile, 4-month-old last-update, missing keywords.

4. 📝

   65 accurate words added, two portals registered

   Django, REST API, Git, Agile added. Naukri score: 84%. NCS portal and T-PRIDE Telangana registered in one evening. Two weeks of 15-minute nightly interview-English practice.

5. ₹

   February 2026 — Offer: ₹3.2 lakh, Madhapur

   Call through updated Naukri profile. Junior software developer. Spoken-English script delivered. First offer in 11 months, after 413 applications.

⚠️ The English-medium gap

There is a second problem, which is harder to talk about because it touches something more personal than portal navigation.

Arjun's medium of instruction through school was Telugu. His college classes were nominally in English but practically in a mixture — teachers who defaulted to Telugu when explaining complex concepts, textbooks in English, examinations in English, lab sessions in both. His spoken English is functional for purchasing and directions but not for the specific register of an IT interview, which has its own vocabulary: "walk me through your approach," "can you explain your data structures decision," "what would you do differently," "tell me about a challenge you faced and how you resolved it." These are not hard questions. They are not testing intelligence. They are testing familiarity with a specific script that is taught, through osmosis, in English-medium schools and in the better-resourced engineering colleges, and that Arjun had never encountered systematically.

In every one of his seven interviews, he said, the point where he felt the conversation turn against him was the spoken English section. He had answered the technical questions — the SQL joins, the Python syntax, the basic algorithm questions — with what he believed was correctness. He had stumbled on the spoken presentation. He had said "I did study about this" instead of "I have studied this" twice in one interview and seen the interviewer's expression shift in the way that expressions shift when a decision has been made. He told nobody in his family about this. His father would not have understood the distinction. His mother would have worried. He had gone home, opened the notebook, and added one more tick to the rejection column.

🌗 What Deepak did

Deepak Rao had gotten his job through his maternal uncle, who worked at a Hyderabad IT services firm and had mentioned Deepak's name to the HR. Deepak knew this. He knew, with the uncomfortable clarity that the beneficiaries of referral economies often have, that his resume was not meaningfully better than Arjun's, that his coding was not significantly stronger, that the difference between getting a job in month three and not having one in month eleven was not ability but access. He had been feeling this knowledge for six months when he visited Arjun in January and found him sitting at the two-chair desk in his Kukatpally room, updating his notebook.

He had been using a Hindi-language AI assistant for about three months for his own work — mostly reading HR policy documents at his company and drafting emails he wasn't confident about. He opened it on Arjun's phone, switched it to Telugu, uploaded Arjun's resume as a PDF, and typed: "Ee resume Naukri algorithm ki ela kanapistundi? Mee opinion cheppagalara?" — How does this resume look to the Naukri algorithm? Can you give your opinion?

The agent read the resume. It replied in Telugu:

"Arjun-garu, meeru chesinadi nijayiti honest resume. Kaani idi Naukri ranking lo okka samasya create avasthundi: last update enta kaalamainadi — November 2024. Daaniki resume rankings 6000+ tarvata padipotundi. Rendu, 'Python', 'SQL', 'HTML' anedi basic keywords — 'Django', 'REST API', 'Git', 'Agile' vantivi add cheyyakkunte MNC-grade JDs keyword match avvadam kashtam. Meeru Nizamabad textile shop ki chesina inventory project lo Django use chesara? Chesaaru ante adi cheppandi — cheppaledu ante add cheyavachu.

Okka vishayam: NCS portal (ncs.gov.in) lo registered unnaraa? T-PRIDE Telangana lo registered unnaraa? Rendu free, rendu government job alerts vastaayi, rendu account Naukri tho link avutundi. Chestara?"

(Arjun-garu, what you have made is an honest resume. But it creates one problem in Naukri ranking: how long since last update — November 2024. Because of that resume ranking falls below 6000+. Second, 'Python', 'SQL', 'HTML' are basic keywords — without adding 'Django', 'REST API', 'Git', 'Agile', matching MNC-grade JDs is difficult. Your inventory project for the Nizamabad textile shop — did you use Django? If yes say so — if not you can still add it. One thing: are you registered on NCS portal (ncs.gov.in)? Registered in T-PRIDE Telangana? Both free, both send government job alerts, both link accounts with Naukri. Shall we do it?)

Arjun read it twice. Then he said, in Telugu, "Django use chesa — kaani cheppalisin antav anipinchindi" — I did use Django but I felt I shouldn't say it. The agent asked why. He said he had thought the project was too small to mention a framework. The agent said the project's scale didn't matter; the keyword did.

"Django use chesa — kaani cheppalisin antav anipinchindi."

— I did use Django, but I felt I should not say it.

📋

Before: honest but invisible

Naukri score: 61%

BSc CS, NPTEL Python, library project, inventory tracker. Skills: Python, SQL, HTML, CSS, basic React. Last updated August 2024. Ranked below 6,000 resumes every week.

🎯

After: honest and visible

Naukri score: 84%

Same projects, 65 words added: Django (already used), REST API, Git, Agile. One-paragraph summary in JD language. Updated that week. Ranked near top for ₹2.5–4L junior roles.

📞

NCS and T-PRIDE: free, one evening

40 minutes total

NCS portal (ncs.gov.in): job matching, career counselling, PMKVY links. T-PRIDE Telangana: quarterly IT job fairs at HICC, walk-in drives. Both free, both registered the same night.

🧭 The three things that changed

Over the next three weeks, three things changed.

The first was the resume. Arjun updated it — added Django, REST API, Git, refreshed the project descriptions to include the specific technical choices he had actually made, updated the last-touched date. He also added, on the agent's suggestion, a one-paragraph summary at the top that used the specific language of the job descriptions he was targeting. He did not fabricate anything. He did not claim experience he didn't have. He added approximately sixty-five words of accurate self-description that he had previously not included because he hadn't known they were the words the algorithm was looking for. His Naukri profile score went from 61% to 84%. The agent explained, in Telugu, what each of those percentage points meant.

The second was the NCS portal and T-PRIDE. He had not heard of either. The NCS portal — National Career Service, run by the Ministry of Labour and Employment — has a Hyderabad regional node with a job-matching function, a career counselling request form, and a skills-certification section that maps to PMKVY courses. T-PRIDE is the Telangana government's placement and recruitment initiative for IT and electronics sector job seekers — it runs quarterly job fairs at Hyderabad's HICC and periodically sends registered candidates to walk-in drives at Telangana-based firms. Arjun registered on both in the same evening, which took forty minutes. The agent walked him through both portals in Telugu, field by field.

The third was the interview English. This was the harder one. The agent had, during one of their practice sessions, asked Arjun to describe the inventory project aloud, in English, as if in an interview. He had. The agent had responded — gently, in Telugu — with a list of specific phrasings he had used that a hiring manager would flag as non-native: the tense inconsistencies, the word-order inversions that come from translating directly from Telugu grammar. It did not mock him. It did not suggest his English was bad. It said: "Ee phrasing correct kaadu antanna kaadu — kaani IT interview lo interviewer oka specific pattern expect chestaru. Aa pattern practise chesthe meeru already cheppina answer baavundi avutundi." — I'm not saying this phrasing is wrong — but in an IT interview the interviewer expects a specific pattern. If you practise that pattern, the answer you already gave will sound strong.

They practised for two weeks, fifteen minutes a night, in the Kukatpally room after dinner. The agent asked him standard HR questions in Telugu and had him answer in English, then pointed to the specific words to change.

The outcome

In February, Arjun got a call from a Hyderabad IT services company — one of the medium-tier ones, not a name you would know, but a real company with a Madhapur office and a proper payroll — for a junior software developer role at ₹3.2 lakh per annum. The call had come through the updated Naukri profile. He had gone for the interview. He had answered the technical questions correctly, as he had before. He had answered the spoken English section with the specific pattern the agent had helped him practise, which is to say he had told the story of the inventory project clearly and sequentially, without tense inconsistencies, using the phrase "what I found was" instead of "what I was finding" and "I chose Django because" instead of "I did choose Django because I thought." These are small changes. They are not the difference between intelligence and stupidity. They are the difference between sounding like you belong in the room and sounding like you don't, in a hiring process that has thirty minutes and three hundred resumes and is making fast subconscious decisions.

He got the offer on a Thursday. He called his father. His father said "Chala santosham" and then, because he is a police sub-inspector with a sub-inspector's composure, not much more. His mother cried. Divya asked immediately what the package was and whether it was more than their father's salary, which is not a rude question in a household where these comparisons are made plainly.

He is still in the Kukatpally room for now. He will move to Kondapur, closer to the office, in June. He is sending Divya ₹2,000 a month for EAMCET coaching materials, which is a thing he had decided he would do when he got a job, which he did not announce to anyone because it was private.

🌱 What a cousin cannot manufacture

Arjun's resume has sixty-five more words than it did in January. They are not new words. Django was already in the codebase. The REST API was already built. The inventory tracker for the Nizamabad textile shop already existed. He had just, for ten months, assumed that the way you wrote a resume was honest and spare, and that the algorithm would see through the words to the work. It does not. The algorithm sees the words.

He moved to Kondapur in June. Divya got into a college in Hyderabad on EAMCET. He sends her two thousand rupees a month for coaching materials. He has not told his father, because in his father's household, the right things are done without announcement.

Deepak's maternal uncle made one phone call. That call was not impressive technology either. It was access — the particular access of knowing what words to put on a resume, what portals to register on, what specific English phrases a hiring manager expects. For most BSc graduates in Kukatpally, that uncle is not available on a Saturday evening in January. The portals are there. The job fairs happen. But none of them explain themselves in Telugu, at eleven PM, in the register of someone who has been through the process.

That is what Deepak installed on Arjun's phone.