BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//sf26-agenda//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:SESS-123@sf26-agenda
DTSTAMP:20261020T151709Z
DTSTART:20261105T193000Z
DTEND:20261105T201500Z
SUMMARY:Fork the Stream: Safe\, Isolated AI Agents on Real-Time Data with A
 gileLog
LOCATION:Room 2012
DESCRIPTION:Shreesha Gopalakrishna Bhat\n\nAI agents are starting to intera
 ct directly with streaming data: querying live topics\, generating test ev
 ents\, building stream processors\, and even writing records back into rea
 l-time pipelines. That is exciting\, but it is also dangerous. Agents are 
 exploratory\, expensive\, and sometimes wrong. If they run against the sam
 e production streams as regular applications\, they can create latency spi
 kes\, interfere with critical workloads\, or publish bad records that down
 stream consumers immediately act on. This talk introduces AgileLog\, a for
 kable shared-log abstraction for making agentic workloads safer on streami
 ng systems. The core idea is simple: give each agent an isolated fork of t
 he stream. Unlike a point-in-time snapshot\, AgileLog’s continuous forks
  keep receiving new records from the parent stream while keeping the agent
 ’s own writes private. Agents can analyze fresh data\, inject synthetic 
 events\, test stream processors\, validate proposed writes\, and promote o
 nly safe results back to the main log. We’ll cover the design of continu
 ous forks\, severed forks\, squash\, and promote; explain how Bolt impleme
 nts these ideas efficiently using a diskless log architecture; and walk th
 rough real agentic streaming examples including IoT analytics\, stream-pro
 cessor testing\, and supply-chain automation. Attendees will leave with a 
 practical framework for thinking about safe AI agents on streaming data\, 
 why existing log abstractions need to evolve\, and what future streaming p
 latforms may need to support as agentic applications move into production.
URL:https://sf26.rmoff.net/s/fork-the-stream-safe-isolated-ai-agents-on-rea
 l-time-data-123
END:VEVENT
END:VCALENDAR