The US edition of this report stands on four federal datasets: a task-level occupation database (O*NET), official wage records and ten-year employment projections (BLS), a tax-record earnings dataset for training programs (College Scorecard), and a national apprenticeship registry. India does not publish equivalents for most of that stack. This edition is built on what India does publish — the PLFS labour survey, the ILO’s occupation-exposure research, the Anthropic Economic Index’s India profile, and government fee and stipend pages — and it is visibly thinner than the US report as a result. Three specific downgrades are marked in boxes like this one where they bite (D1: no earnings-by-program data; D2: exposure carried across taxonomies; D3: prices assembled page by page). Where a number cannot be printed honestly, you will see the gap stated instead of a number. That is the product.
You answer customer calls, chats and emails for a company or its outsourcing partner: product questions, complaints, account changes, basic troubleshooting — scripted where possible, escalated where not. In India this is the core of the voice- and chat-BPO industry.
Occupation identified as Contact Centre Information Clerks — ISCO-08 unit group 4222, which is also the NCO-2015 unit group (India’s classification shares its first four digits with ISCO-08)[S1].
No wage for your specific occupation. PLFS publishes earnings by employment status and sex, not by detailed occupation. Occupation-level medians (your unit group, your state) exist only inside the PLFS unit-level microdata — free, but behind a one-time registration this build does not hold. The moment that registration exists, this block upgrades from “all salaried workers” to “your occupation, your state”[S4].
No official employment projections. The US report prints the government’s own 10-year outlook per occupation. India publishes none — no MoSPI or NCS dataset projects occupation-level employment. Where the US edition shows a projection, this edition shows this sentence. Vacancy listings on the National Career Service are the nearest official signal, and they are listings, not projections[S5].
Index one — the ILO’s refined GenAI exposure index (the lead index of this edition, chosen because it is built on the same ISCO-08 taxonomy India’s NCO-2015 aligns to). It scores the technical possibility that the tasks of an occupation could be performed with GenAI — assembled from nearly 30,000 task ratings. The authors’ own reading instruction, verbatim:
“[W]e stress that such exposure does not imply the immediate automation of an entire occupation, but rather the potential for a large share of its current tasks to be performed using this technology. Whether this leads to the disappearance of an occupation or workforce replacement is a more complex question — one that will depend on the initial decision to adopt the technology, but also the extent to which individuals in these occupations are given opportunities to learn to work with these technologies and adapt to the evolving nature of their tasks.”[S2]
Outbound sales-side contact-centre work is a different unit group — 5244 Contact Centre Salespersons — and scores HIGHER (Gradient 4, 0.61). If your work is mostly selling on calls, read this report against 5244, not 4222; both rows print in Section 2.
One classification honesty note: this report reads your job against ISCO-08's definition of 4222 (inbound contact-centre information and service work). NCO-2015's own enumerated job titles are placed slightly differently: the literal title 'Customer Service Executive' is listed under 4225 Inquiry Clerks, 'Customer Care Executive (Call Centre)' under 5244 Contact Centre Salespersons, and family 4222 itself lists only the technical-support voice/non-voice roles. All three families print in Section 2 (readings 0.57, 0.61 and 0.58) — every coding of this job lands in the exposed gradients, so the reading does not turn on the coding choice. We show the quirk rather than smooth it.[S1]
Index two — observed AI usage, India profile. In the Anthropic Economic Index (period 2026-05-01), India’s share of observed Claude usage is about 0.3× what its working-age population share would predict — rank 102 of 121 covered countries. Inside that usage, the learning skew is the story: Education & Learning requests are 19.3% of Indian usage against a 13.2% global baseline, and coursework use runs 19.9% vs 16.4%. Automation-style usage (51.9%) slightly exceeds augmentation (48.1%) in India — the reverse of the global split[S3].
For your occupation specifically, the honest line is: AEI publishes no India row for customer-service-representative tasks this release — the data fell below the index’s privacy threshold. Per the publisher, null means not published, never a measured zero. The nearest published India rows in the contact-work family: tasks of Telemarketers, 0.53% of Indian usage; General Office Clerks, 0.45%; Office & Administrative Support as a category, 7.5% (global baseline 7.9%)[S3].
The dataset’s own publisher, verbatim: “This data alone cannot support conclusions about job displacement or job security, and cannot ground advice on career choices in either direction.” And the accurate frame is “AI is used for tasks commonly done by [occupation]” — not “[occupation]s are using AI.”[S3]
Maharashtra is rank 1 of 27 published Indian subregions in the AEI by observed usage (the index publishes a rank for Indian states, not a usage-share percentage — so no percentage is printed here). Work-related use runs 45.5% of Maharashtra’s usage, and Office & Administrative Support tasks are 7.8% of it — your occupational family is measurably present in the state’s AI usage[S3]. Route prices in Section 3 are Maharashtra government figures where marked; the state’s own occupation-level wage medians join this report when the PLFS microdata registration exists (Section 5).
The US edition reads exposure from indices built natively on the US O*NET/SOC occupation system. No India-native task-level exposure index exists. This edition’s lead index (ILO WP140) is built on ISCO-08 — which NCO-2015 shares its first four digits with, so the join is a single, official alignment, not a guess. But the ILO itself warns that the task mix inside the “same” ISCO occupation differs across countries: an Indian contact-centre job is not task-identical to the global average the score describes[S2].
Two US indices were deliberately NOT carried over: the Felten AIOE score would need a double crosswalk (SOC→ISCO→NCO) whose official bridge file (bls.gov) refused this build’s fetches, so no AIOE number prints anywhere in this report; and AEI occupation rows are US-SOC-framed, so they print above only as India-published usage shares with that framing named. A report that quoted those numbers without the crosswalk would look thicker and be less honest.
The US edition proves, by recomputation on the U.S. Labor Department’s skills graph, that “safe adjacent pivot” lists are mostly fiction (correlation r = 0.907 between an occupation’s exposure and its skill-neighbours’). India publishes no related-occupations graph for NCO-2015, so that computation cannot be run for India — and this report says so instead of pretending. What can be shown, from the ILO table directly, is the same pattern at family level: the occupations nearest to contact-centre work are all in the exposed gradients.
| Clerical & contact-family occupations (ILO Table A1) | ISCO-08 / NCO-2015 group | ILO exposure |
|---|---|---|
| Contact Centre Information Clerks you are here | 4222 | Gradient 3 · 0.58 |
| Data Entry Clerks — Data Entry Operator | 4132 | Gradient 4 · 0.70 |
| Typists and Word Processing Operators | 4131 | Gradient 4 · 0.65 |
| General Office Clerks — Office Clerk | 4110 | Gradient 4 · 0.60 |
| Contact Centre Salespersons — Telesales Executive | 5244 | Gradient 4 · 0.61 |
| Inquiry Clerks — Enquiry Clerk | 4225 | Gradient 3 · 0.57 |
| Receptionists (general) — Receptionist | 4226 | Gradient 3 · 0.57 |
| Office Supervisors — Team Leader (BPO) | 3341 | Gradient 2 · 0.43 |
Exposure tiers and means extracted from ILO WP140 Table A1 at build time[S2]. This is a family-level reading from the index itself — NOT a skills-adjacency computation, because the data to run one does not exist for India. The reading instruction is unchanged: nearby desk work mostly shares your exposure; the moves that change the number change the work, and those moves are priced in Section 3.
Of the thirteen occupations in the ILO’s highest exposure tier, nine are clerical or contact work — the family your job sits in. Moving one desk sideways keeps you inside that exposed family (data entry 0.70, typists 0.65, general office 0.60, contact-centre sales 0.61, inquiry desk 0.57, reception 0.57, supervision 0.43 — against your 0.58; every one of these sits in the exposed gradients). The destinations that carry materially lower readings — IT support (0.47, high variability), supervision (0.43), electrician (0.19), healthcare support (0.14) — all involve new skills, which means retraining, which has a price. In India that price runs from ₹0 to about ₹27,000 on government routes. That is what Section 3 prices.
Below are the moves that actually change the number — with real institutions, published fees in ₹, real durations, and the exposure reading of each destination. Prices are catalog facts as of the date shown; institutions change them at will, so treat every figure as “verify before you enrol.” Each price carries a verification tag: green means the build fetched the institution’s own page that day; amber means a named aggregator supplied it and the institution’s page must be checked before you pay anyone.
The US edition prices programs from a single federal dataset with a price field per program. India has no such dataset: government-ITI fees are set state by state, NIELIT fees are set centre by centre, and scheme funding changes batch by batch. Every price below was therefore assembled from an individual institution or scheme page, each with its own URL, check date and verification tag — and the ITI figures are Maharashtra’s: another state’s reader needs another state’s page. Fewer routes are priced here than in the US edition for exactly this reason; a route without a checkable price prints as a pointer, not a number.
Before you pay anyone: the Government of India runs the National Career Service (ncs.gov.in, Ministry of Labour & Employment) — free job matching, free career counselling, and listings of skill courses and apprenticeships. Its services are free of cost. It will not price routes for you the way this report does, but every route below can be cross-checked against it, and registering costs nothing.
National Career Service: ncs.gov.in[S5] · free-course catalog: skillindiadigital.gov.in (browsing is public; enrolment is Aadhaar-gated) · credential check: nqr.gov.in — The National Qualifications Register is the official register of NSQF-aligned qualifications. A credential that appears there is government-recognised at a stated NSQF level; one that does not, is not. Check any certificate a seller waves at you against it.
The nearest real move that changes the work: from scripted voice support to hands-on IT support. Customer-handling under pressure transfers directly; what the credential adds is the technical half — systems, networks, troubleshooting. NIELIT is a government institute under the Ministry of Electronics & IT, and its 'O' Level is the long-standing entry credential for government and private IT support roles.
ILO WP140 scores this destination Gradient 2 (mean 0.47) — but with the highest task-variability of any destination on this menu (SD 0.19). Read that honestly: parts of IT support work are themselves GenAI-exposed; the destination is lower-exposed than contact-centre work, not safe.
The full exit from screen-and-script work into a hands-on licensed trade. Nothing clerical transfers except discipline; that is exactly why the exposure number drops so far. Two years at a government ITI in Maharashtra costs less than most private 'AI-proofing' webinars — the constraint is seats and the work is physical.
ILO WP140 places electricians (7411) in its lowest tier — Not Exposed, mean 0.19: 'most tasks remain relatively unaffected by GenAI.' Electrical fitters (7412, mean 0.17) read the same.
India's no-tuition-first route: apprentices are engaged by an establishment and paid a government-notified stipend from month one, across the scheme's 49 sectors — search the portal for openings in your target field. For a BPO worker the honest use of this menu item is paired with a trade or credential above — apprenticeship is how the first year of a new field pays you instead of you paying it.
Where sources disagree, we show it: Both government sources were fetched at this build (PIB via browser-UA fetch after an earlier 403). They agree on the band (₹6,800–12,300) and that it is NOT yet in force (AIR: 'recommended… once notified'; PIB: 'proposed'). They disagree on the raise percentage — AIR's headline says 30%, PIB says 36% (₹5,000→₹6,800 is arithmetically +36%) — shown here rather than smoothed.
The service half of BPO work — patience, difficult conversations, process discipline — is the transferable core of patient-support work. The clinical half is what a GDA course teaches. The lowest exposure reading on this menu, and the route where India's free government skilling catalog actually reaches.
ILO WP140: Health Care Assistants (5321) — Not Exposed, mean 0.14, the lowest reading on this menu. Associate-level nursing (3221, Minimal Exposure, 0.22) is the licensed rung above it, via ANM/GNM programs — a longer, regulated path not priced here.
If you stay in contact-centre work, staying should be a decision with a price tag too. India is the one market where the bottom rung of that ladder is nearly free: government MOOC platforms charge nothing to learn and ₹1,000 to certify. The honest caveat is the same as the US edition's: no published evidence ties any certificate to job retention. Both facts belong in the same sentence.
One rung verified from the government platform's own page at this build. Vendor AI certificates (the $99–$500 US band) exist in India at import prices — they are not listed here because a ₹-priced government rung exists and the evidence for either improving job retention is equally unpublished.
The US edition prints, for many programs, the actual median earnings of past completers, from tax records (College Scorecard). India publishes no earnings-by-program record — none. The closest thing that exists is NIRF, where ranked institutions self-report the median salary of their placed graduates: self-reported, placed-graduates-only, ranked-institutions-only, lightly audited. None of the government routes on this menu lands at a NIRF-ranked institution, so no earnings figure appears on this menu at all — and any competing report that shows you one is quoting either an institution’s own marketing or an unsourced estimate. Where this report shows a salary figure it is the institution’s own NIRF submission, so labeled; where it shows none, none exists[S10].
| Destination | ILO exposure | Note |
|---|---|---|
| Team Leader / Office Supervisor (promotion in place) | Gradient 2 · 0.43 | the internal ladder out of the phone queue; exposure drops to Gradient 2 (0.43) — a real but partial move, and openings are the employer's to give |
| Technical support (non-voice IT helpdesk) | Gradient 2 · 0.47 | same family as the NIELIT route above; some BPO employers cross-train internally — ask before paying for training they may fund |
| ANM/GNM nursing (licensed) | Minimal Exposure · 0.22 | the regulated rung above GDA: 2–3.5-year state-regulated programs; fees vary by state and institution — priced per institution, not printable as one number |
Every figure in this report carries a bracketed marker resolving below. Access dates are the dates each source was actually fetched — by this build where marked, otherwise by the venture’s dated India data screen.