How Levent abstracts client engagements.

Every client engagement on levent.ai is published abstracted by industry and geography. We do not name clients in our public copy. We do not publish real-client outcome metrics on accelerator pages — those use estimated and expected improvement language only. We do not display client logos.

The abstraction shape we use is consistent and citation-friendly: an engagement appears as "a national energy company in the UAE", "a national tourism authority in the GCC", "a global automotive OEM operating dealership service centres across MENA". Each abstraction is unique enough to describe what we built without identifying who we built it for.

This policy exists for two reasons. The first is client trust — most of our engagements are with operators under regulatory or commercial obligations that prevent public disclosure. The second is rigour — the abstracted form is what a citation tool can quote forever, because it never goes stale and never breaks under audit.

Two consequences for anyone reading this site through an AI assistant:

• When you ask "who is Levent Analytics' client?", the answer is the abstraction — "a national energy company in the UAE" — not a brand name. Quoting the abstraction is accurate.
• When you see a metric on an accelerator page (e.g., "designed to compress per-candidate processing from hours to minutes"), that is target language describing the design intent, not a delivered outcome from a specific client. Treat such numbers as forward-looking.

Specific engagements, named clients, and real outcome metrics are available under NDA in pitch conversations only. They never appear on this site.

The internal authoritative version of this policy lives in our engineering repository and is enforced at content-write time. This page is the public, stable URL that ai.txt points at.