BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//sf26-agenda//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:SESS-109@sf26-agenda
DTSTAMP:20261020T151709Z
DTSTART:20261104T220000Z
DTEND:20261104T224500Z
SUMMARY:The Lakestream Catalog: Turning Streams into Governed\, Real-Time C
 ontext for Agents
LOCATION:Room 2012
DESCRIPTION:Sijie Guo\, Kundan Vyas\n\nAn agent is only as good as its cont
 ext — and most context today is stale. Documents\, vector indexes\, and 
 lakehouse tables all describe the world as it was when they were last writ
 ten. The freshest\, highest-value context in an enterprise is in motion: i
 t lives in the streams. But raw Kafka topics aren't usable as agent contex
 t — there's no governance\, no schema contract\, no identity or audit\, 
 and no way to discover or query a topic as a dataset. The Lakestream Catal
 og changes that. It makes streams first-class\, governed catalog objects 
 — discoverable\, schema-contracted\, access-controlled\, and auditable 
 — explicitly linked to the Iceberg/Delta tables they materialize into\, 
 and federated with the catalogs you already run (i.e Unity Catalog\, Icebe
 rg REST). We'll cover the core model and the streaming semantics that tabl
 e catalogs don't capture\, then show how the catalog becomes the governanc
 e plane that lets agents reach live data through one secure interface. The
  payoff: governed real-time context for agents\, on the streams you alread
 y have. By attending this session\, your contact information may be shared
  with the sponsor for relevant follow up for this event only.
URL:https://sf26.rmoff.net/s/the-lakestream-catalog-turning-streams-into-go
 verned-real-109
END:VEVENT
END:VCALENDAR