Privacy choices
This site may send the owner a minimal page-visit notice with page, browser, device type, and resolution. If you accept analytics, it can also remember an anonymous visit/session ID and collect richer engagement signals.
Read the privacy noticeOptional Postgres extension Alex OS uses for content embeddings, with an automatic keyword-scoring fallback when it's not configured.
Used only inside Alex OS's vectordb service, not in the CMS. When database_url is unset or unreachable, retrieval falls back to keyword scoring instead of failing.
Alex OS now supports pgvector for semantic search while staying fully functional without it — keyword scoring kicks in automatically when the extension isn't c…
What changed: optional embeddings, guaranteed retrieval
pgvector is wired into Alex OS's vectordb service as an opt-in Postgres extension. When `database_url` is set and reachable, content embeddings power semantic search. When it's not — missing config, unreachable host — the system falls back to keyword scoring automatically. No crash, no manual intervention. The CMS layer is untouched; this lives entirely inside the vectordb service.
Why this matters for the build
The fallback design means you can run Alex OS locally or in a stripped-down environment without provisioning a Postgres instance at all. When you're ready to upgrade to full semantic retrieval, you configure `database_url` and pgvector takes over. If you're working with the embeddings backend directly — say, forcing a reindex after a raw DB write — the reindex guide targets exactly this layer.