Vector Database
Vector Database explained
The database supplies search infrastructure; it is not the language model. Retrieving document passages, generating an answer and checking its evidence are different tasks. Separating them helps diagnose problems: did retrieval find the wrong passage, or did generation misrepresent the right source?
Selection depends on collection size, updates, filters, access control and operations. A small internal asset collection differs from a large product catalogue. Start from those requirements; blanket cost comparisons between specialist and integrated systems do not provide a sound basis.
pgvector is a documented example of vector search within PostgreSQL. Depending on the method, a search can determine nearest vectors exactly or use approximation to favour speed. Technical neighbour accuracy differs from domain relevance: precisely similar vectors may still represent unsuitable documents.
Changed and deleted content needs a clear route into the live search collection. Permissions must take effect in actual queries. A successful demonstration does not establish that obsolete price lists, withdrawn assets or another project’s files will be excluded reliably.
Examples
Hypothetical application
An tests retrieval over approved product documents. Alongside speed and relevance, it checks that a withdrawn version disappears and each team can retrieve only its permitted documents.
Key Points
- Separate retrieval infrastructure from answer generation.
- Assess domain relevance separately from nearest-neighbour accuracy.
- Test updates, deletion and permissions.
Practical application
Test selection against a realistic, approved collection, including and permission changes.
Useful measures
Relevant results
Assessment against documented search tasks.
Update delay
Time between a content change and its correct reflection in search.
Common mistakes
- Using database speed as evidence of answer quality.
- Leaving deleted or restricted content retrievable.
Sources and context
- pgvector: Vector similarity search
Official documentation for vector search in PostgreSQL.
Frequently Asked Questions about Vector Database
No. Many applications need no vector search. The task determines the requirement.
It searches faster but may miss some mathematically nearest items. Evaluate the trade-off with the actual collection and tasks.
There is no sound answer without requirements and usage. Compare setup, operations, maintenance, search load and required quality.
Loading related terms…
All TermsArticles about Vector Database
AI with Brand DNA: Why Generic Bots Are a Brand Risk
Off-the-shelf AI assistants can dilute your brand. Learn how strategic calibration, guardrails, and red-teaming can transform a generic bot into a powerful, on-brand ambassador that positively impacts business outcomes.
Grok Bot Skills: How to Truly Scale AI in Your Marketing
Reusable task instructions help teams organise AI work consistently. Learn how to connect briefs, data, quality checks and version control, and measure the benefits in your own workflow.
Prompt Ops: The Operating System for AI in Marketing
The uncontrolled use of AI prompts leads to chaos. Prompt Ops provides a structured approach to manage prompts like software, ensuring efficiency, quality, and scalability in marketing.