Key takeaways
- White-collar hiring rose 14% year on year in October 2026. The Naukri JobSpeak index was 2,836, up from 2,480 in October 2025, as reported on 3 November. Naukri itself warns that part of this comes from a weak base: Diwali fell in October last year.
- AI/ML roles grew 32%. "IT and software services" grew only 5%, and software engineering and QA roles across industries grew just 1%. The money is going to AI work and to companies that build products for their own use, not to generic coding headcount.
- Global Capability Centres (GCCs) grew 18%, BPO/ITeS 17% and banking and financial services 16%. Inside BPO/ITeS, software engineering and QA roles grew 42%, as reported. Tech roles are spreading outside "IT companies".
- Freshers (0–3 years) saw 15% growth, a little ahead of the 4–7 year band (11%). Senior profiles (13+ years) grew the most, at 22% (SIA).
- Hyderabad (+15%) and Pune (+14%) led the metros. Bengaluru grew 12%, and Gujarat cities such as Gandhinagar (+48%) and Surat (+36%) grew fastest from small bases.
- What to do: stop applying with a plain "Java/Python developer" profile. Add one AI-adjacent skill you can show working (RAG, evaluation, data pipelines, MLOps), and target GCCs, BFSI and ITeS tech teams as seriously as the big IT brands.
1. What the October JobSpeak actually says
Naukri JobSpeak is a monthly index built from new job postings on Naukri.com. It measures hiring demand, not hires made. The October 2026 edition came out on 3 November. These figures come from coverage by Lapaas Voice, Staffing Industry Analysts, SightsIn Plus and CareerIndia. The headline numbers are the same across all of them.
Two caveats before you read too much into a 14% jump:
- Base effect. Info Edge's CEO said that some of October's growth reflects last year's festive-season base. In September 2026, before that effect, the index was up only 2% year on year (Naukri blog, September edition).
- Month on month, the index fell, from 3,053 in September to 2,836 in October, as reported. That is normal in a festive month. It means "the market is improving", not "the market is booming".
The more useful signal is the mix. One category has now grown strongly two months in a row: AI/ML roles were up about 20% in September (as reported) and up 32% in October.
2. Where tech hiring grew and where it didn't
| Segment (October 2026, YoY) | Growth | What it means for a tech jobseeker |
|---|---|---|
| AI/ML roles | +32% | Fastest-growing tech category; still a smaller absolute pool than general software |
| GCCs | +18% | In-house tech centres of global firms: product, platform, data, security |
| BPO/ITeS overall | +17% | Its software engineering and QA roles grew 42%: automation and tooling teams |
| Banking & financial services | +16% | Fintech, core-banking modernisation, risk and fraud data teams |
| Freshers, 0–3 yrs (all sectors) | +15% | The first rung is not shrinking in this data |
| IT & software services (industry) | +5% | Services firms hiring, but slowly |
| Software engineering & QA (function, all industries) | +1% | Generic dev/QA demand is almost flat |
Sources: JobSpeak October 2026 as reported by Lapaas Voice and SIA.
Read the first and last rows together: AI/ML +32% vs software engineering and QA +1%. Companies are still hiring engineers. They are just labelling and scoping the roles differently. A posting that asks for "backend engineer, experience with LLM APIs and vector search" is increasingly filed as AI/ML work. Its core is still backend engineering.
This is consistent with Naukri's July World AI Day report. It found AI-related skills in 15.4% of technology job descriptions in 2026, against 0.9% in 2020. It also said recruiter demand rose in the ₹40–49 lakh and ₹50 lakh-plus salary brackets. In other words, AI skills are becoming a line item in ordinary engineering jobs, and the premium sits at the experienced end.
3. Cities: the map is widening
| City | October 2026 YoY |
|---|---|
| Hyderabad | +15% |
| Pune | +14% |
| Delhi-NCR | +13% |
| Bengaluru | +12% |
| Mumbai | +12% |
| Chennai | +11% |
| Kolkata | flat |
| Gandhinagar / Surat / Vadodara / Ahmedabad | +48% / +36% / +30% / +23% |
Note: the Gujarat figures cover all white-collar roles, not only tech, and they start from small numbers. Manufacturing roles grew strongly in the same month, so don't read these numbers as a tech boom.
For Noida and Gurugram engineers, NCR's 13% is close to the national average. Hyderabad and Pune are the two metros where hiring is growing faster than the national rate. Hyderabad in particular has a dense cluster of GCCs. If you are open to moving, those two cities are where a sharpened profile gets the most interviews per application.
4. What this means by profile
| You are… | Market reality from this data | Do this |
|---|---|---|
| Final-year student / 0–1 yr | 0–3 yr demand +15%, but generic dev/QA nearly flat | One solid project that uses an LLM API properly (retrieval, evaluation, cost logging), deployed and documented |
| Services developer, 2–5 yrs | Your industry grew 5%; GCC and BFSI tech grew 16–18% | Rewrite your CV around systems you ran and numbers you moved; apply to GCCs and bank tech centres |
| Manual / functional QA | Function +1% overall, but +42% inside BPO/ITeS automation teams | Move to test automation (Playwright, API testing) plus AI-output evaluation |
| Data / analytics, 2–6 yrs | AI/ML +32% is the closest adjacent pool | Add MLOps basics: pipelines, model serving, monitoring, data quality checks |
| Infra / cloud / NOC | Not broken out, but AI workloads need GPUs, storage and networks | Learn GPU node ops, Kubernetes and cost control; this is what "AI infra" job titles ask for |
5. A 60-day plan that fits the growth areas
Days 1–15: reposition, don't restart. Take your current stack and add one AI-facing capability on top of it. A Java developer builds a Spring service that calls an LLM with retrieval over company-style documents. A QA engineer builds a test suite that scores LLM answers against expected outputs. Don't start from scratch in a new language.
Days 16–35: build one thing that runs. Deploy it on a free or low-cost tier. The cheapest Indian VPS plans or a cloud free tier are enough for a portfolio. Log latency, token cost in ₹ per 1,000 requests, and failure cases. Recruiters for AI/ML roles screen for exactly this: can you measure and control it, not just call the API?
Days 36–50: make it visible. Write a README with an architecture diagram, numbers and known limits. Put the project link at the top of your Naukri profile and CV. The AI-skills report found AI skills in 15.4% of tech JDs, so recruiters filter on those terms. Use the exact phrases a JD would use: "RAG", "LLM evaluation", "MLOps", "vector database". Use them only where they are true.
Days 51–60: aim the applications. Split them roughly evenly between GCCs, BFSI tech teams and ITeS automation teams, with a smaller share to large IT services firms. Prioritise Hyderabad and Pune if you can relocate.
6. What to watch next
- November JobSpeak (early December): with no festive base effect, it will show whether AI/ML growth holds without help from the comparison.
- Q3 FY27 IT results (January): fresher intake numbers from the large services firms are the clearest test of whether the 0–3 year growth turns into joining dates.
- AI-skill share in JDs: if Naukri's next update moves well above 15.4%, "AI-adjacent" is becoming a baseline requirement, not a differentiator.
The October data is good news with a footnote. Hiring is up, freshers are included, and the growth is concentrated in AI work and in-house tech centres rather than generic coding roles. Position yourself where the growth is: same core skills, one AI capability you can prove, and applications aimed at GCCs, BFSI and the faster-growing cities.
Planning your next move? Read our October hiring analysis for freshers and how to choose an LLM for your product, then compare offers with the salary in-hand calculator and tighten your CV with the resume builder.