About
A dictionary tells you what Kubernetes is. This tells you what to ask next. Every one of the 659 terms is written in plain English, says what it signals about a candidate, and ends in a question you can put straight on a screening call.
It tells you which terms not to ask about
Every term is weighted by whether its presence actually separates one candidate from another. 121 carry real signal and are worth leading a screen with. Many of the rest are table-stakes in their track, near-universal, and asking about those is how a screening call goes wrong. No other glossary warns you off its own content.
It catches impossible experience demands
262 terms carry a checkable release year, so the decoder can show you where a job description asks for eight years of a six-year-old tool. That gives you an argument to take to a hiring manager, rather than a definition. Saying a requirement is unmeetable ends a conversation that "this seems like a lot" does not.
It turns a pasted JD into a screening brief
Paste a job description and it infers the role archetype, ranks the terms that actually differentiate candidates, and writes a screening script you can read down the phone. The brief is built from the same weighting as the glossary, so it leads with what matters and leaves the table-stakes terms out of your mouth.
It tells you when a missing skill is not missing
32 curated clusters of genuinely interchangeable skills mean a candidate who lists Google Cloud instead of Amazon Web Services stops being a false reject. The clusters are deliberately conservative: only substitutes a hiring manager would actually accept, not every tool that shares a category. The same clusters power the JD-against-resume comparison, where the useful answer is almost never the match percentage but the middle category — covered by something else.
It tells you where that person works
The talent map reads the decoded JD and names the pool: 8 talent clusters, the kind of employer to target, 401 curated companies holding a real bench, and which of 12 Indian hub cities they sit in, with what each city is like to hire in. Filter it to product companies when a services pipeline is not what you need. Hand-curated and dated in the tool, last reviewed October 2026 — a shortlist to verify, not a live database.
It writes the Boolean string for you
Four copy-ready searches come out of the same decode: titles plus core skills, a tight skills match, a wider net that folds in the substitutes above, and a target-company search. The OR groups are built from the aliases the glossary already tracks, so Kubernetes carries K8s and Bengaluru carries Bangalore. Plain AND, OR and NOT only, so they paste into LinkedIn, Naukri, an applicant tracking system or a Google X-ray without editing.
It covers the look-alikes
Java against JavaScript, Angular against AngularJS. 54 pairs that read as the same skill on a resume and are not, set side by side with what each one actually implies about a candidate.
It separates the roles that get mis-screened
28 role profiles cover the clusters where two or three job titles get used interchangeably and mean different things. Each lists the terms a credible resume should show, the three whose absence is a genuine red flag, and decoy terms that read as related but do not signal fit. Alongside them, 8 rungs of the engineering ladder explain what each title is expected to own, and why the same title means different things at a startup and a bank.
It gives you the prompts, already written
87 copy-paste AI prompts across 16 categories, from job descriptions and sourcing to compensation and compliance. Each says when to reach for it and how to get a usable answer, which is the part a prompt list normally leaves out.
Who made it
Built by Rajiv. It is an early beta and still growing. If a term you keep meeting is missing, the glossary has a suggestion box on every screen.