VISTAR Is the Data Spine My Job-Matching Grid Needed — Now Use It to Publish Monthly Jobs Data, District by District

"Better data means better decisions, better policies and better delivery to people," the Prime Minister wrote on 4 November 2026, welcoming the Cabinet's approval of VISTAR — as DD India reported the next morning. I have been arguing a narrower version of that sentence for years. The Cabinet has now funded the plumbing: ₹4,371.10 crore, 2026-27 to 2030-31, per PIB. This post asks MoSPI to spend part of it on three specific things, and flags one risk.

I want to be precise about what the PIB release of 4 November 2026 actually promises, because the detail matters more than the outlay. In its own terms:

• 59 surveys in the 16th Finance Commission cycle, against 43 major surveys previously
• AI and machine learning embedded in regular workflows, "primarily for data processing, validation, and analysis"
• harmonised "AI-ready" datasets, built on data cataloguing, machine readability, interoperability and privacy by design, explicitly to "dismantle silos"
• mobile phone, scanner and e-commerce data brought in to supplement surveys
• district-level and million-plus city estimates for PLFS, HCES and ASUSE
• the 8th Economic Census, down to Gram Panchayat and ward level, capturing "workers hired, therein", as the baseline for a National Statistical Business Register

That last pair of bullets is the one I cannot read calmly. In National Job Skilling Policy (April 2026) I quoted the NITI Aayog working paper's demand for an "Annual Skills and Employment Survey + central data portal" feeding "a digital data bank and portal to track demand, supply, and placement outcomes". I did not know a funding vehicle was coming. VISTAR is that vehicle.

So, three requests.

Publish monthly, district by district, on a declared calendar. I made this case in Monthly Jobs Data to be Released from April (March 2025). A district figure that surfaces two years late guides nobody's decision.

Wire PLFS micro-data into the matching engine. In National Employment Grid : a Whitepaper and Bharat Job Alert Mission (both July 2025) I proposed AI matching across eShram, NCS and Udyam. Those registers know who is looking. They do not know where demand is rising. PLFS does.

Make the Economic Census register the employer side of the grid. Enumerate every non-farm establishment to ward level and each becomes a verified, geocoded job poster — including the three-person workshop that will never open a job portal. My Census 2027 post (March 2026) welcomed a digital population census; this is its economic sibling.

Now the caveat I owe you. "AI-ready" must not mean "un-auditable". Reproducible methods, published documentation of what the models do, and an outside check on the outputs are not optional extras. And commercial digital data carries real questions of representativeness and privacy that a press note cannot settle — The Policy Edge made exactly this point on 4 November, noting that bringing such data into official statistics makes safeguards and comparability essential. Access to a dataset is not the same as confidence in it.

Since 2018 I have been asking for net employment growth, industry-wise and region-wise, set against graduation output (#UnEmployment #JobsData #EPFO #DigitalIndia). For the first time, I think a ministry has the money to answer me. I will be watching the first publication date more closely than the outlay.

With regards,
Hemen Parekh
www.Iam-Immortal.ai / www.IndiaAGI.ai / www.YourContentCreator.in
5 November 2026