Lakshya

Method

How this was built, and where it is weak

Stated plainly, because a market map you cannot audit is worth very little.

Collection. Public job-board APIs — Greenhouse, Ashby and Lever — queried directly in August 2026. 222 company boards probed, 122 resolved, 14,011 live postings retrieved. No scraping, no logged-in sources, and nothing behind a terms-of-service prohibition.

Selection. 1,558 postings matched an AI keyword pattern on title; 1,405 matched a platform, SRE, DevEx or SDLC pattern. Full description text was parsed for 592 US and 72 India AI postings plus 427 platform postings, and term frequencies computed by regular expression.

Compensation. Extracted from description body text by identical method for both families. US base salary only — n=1,237 for AI, n=515 for platform. Per-archetype bands are an apportionment of those corpus figures by title and seniority, not directly measured per archetype.

Interview loops come from published 2026 interview guides and candidate reports, not from the posting corpus, and are the least certain content on the site. Treat stage counts as typical rather than exact.

Four limits that matter

  • The India sample is small and unrepresentative. 72 AI roles from 11 boards. Naukri, Workday-hosted employers and companies on their own portals are absent, which excludes most large GCCs and most of the services sector. Treat every India number as directional only.
  • Compensation is biased upward. Companies publish bands mostly because California, New York, Washington and Colorado require it — so the median is a median of the transparent, not of the market.
  • Title matching is imperfect in both directions. Terms like “model” and “agent” pull false positives; real AI work under a plain “Software Engineer” title is missed entirely.
  • Term frequency measures mention, not importance. A posting naming Kubernetes once counts the same as one built around it.

Why not LinkedIn or Naukri

Both prohibit automated collection in their terms, and Indian law is materially less scraper-friendly than US law — the IT Act 2000 §43(b) creates civil liability for copying data from a computer resource without permission, with no requirement to show a technical barrier was defeated.

It is also unnecessary. Both sites are mirrors: the postings originate in Greenhouse, Ashby, Lever and Workday, all of which serve structured data publicly to their own front ends. Reading the source gives cleaner fields than the mirror does.