<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/acff991e3da94d5aa4e98dcee0b100e2&quot; frameborder=&quot;0&quot; width=&quot;1920&quot; height=&quot;1440&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1440</height><width>1920</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1440</thumbnail_height><thumbnail_width>1920</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/acff991e3da94d5aa4e98dcee0b100e2-14851ea806de7898.gif</thumbnail_url><duration>32.36</duration><title>33s RAG walkthrough (pgvector HNSW + Gemini)</title><description>End-to-end RAG demo of Knowlex, the AI knowledge-retrieval half of craftstack.

Shown in 33 seconds:
- /kb ingest — chunk + embed (gemini-embedding-001, 768 dim)
- / ask — cosine kNN over pgvector HNSW, streamed Gemini 2.0 Flash answer with numbered citations
- /api/kb/stats — live index-type probe
- /docs/api — hand-written OpenAPI 3.1

Stack: Next.js 16, Prisma 7, Neon Postgres + pgvector, @ai-sdk/google.
Companion: Boardly walkthrough — https://www.loom.com/share/1f6915e588cb4176bfc8272f0f9310bb
Repo: https://github.com/leagames0221-sys/craftstack</description></oembed>