Semantic search

Show which results your pipeline registered.

Semantic search is harder to audit than keyword search. Results that look right may not be authorized. Mnemo returns a vector_uid when you embed a document and again when a retrieved vector verifies.

01Where it applies

The vector-first filter.

If you embed a corpus and let users query it semantically, Mnemo applies. Enterprise knowledge search, legal research, medical literature, support knowledge bases, document discovery.

02Why it matters

Results should show whether they verify.

The retriever cannot tell a document your pipeline embedded from a lookalike written into the index some other way.
03What this does

Show which results verify.

Every document your pipeline embeds gets a vector_uid registered to your account. Verifying a result at query time reports whether it resolves to an object your pipeline registered. Surfacing only verified results is one conditional in your retriever.

04Scope

What Mnemo does not do.

Does not replace your system
Does not tune relevance
Does not moderate content
Does not score document quality

Mnemo focuses on one layer: attaching identity to embeddings and verifying them when they are used.

Mnemo Verify checks whether a submitted vector resolves to an object registered to the caller's Mnemo account or namespace. verified=false is a safe outcome, not an error, and not proof that a vector is foreign. Verify does not establish the truth, authorship or ownership of the content behind a vector.

05Who this is for

CTO at search-heavy products in legal, medical, and enterprise knowledge. Head of Search Engineering. Head of Trust and Safety.

06Why teams adopt it
  • Enterprise search teams need to show which indexed vectors came through their own pipeline.
  • Customer-facing search in regulated industries needs citation evidence.

Start verifying in minutes.

Start with a 30-day free trial.