BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//open-source-conference-luxembourg-2026//ta
 lk//ZZV37G
BEGIN:VTIMEZONE
TZID:Europe/Luxembourg
BEGIN:DAYLIGHT
DTSTART:20251007T000000
TZNAME:CEST
TZOFFSETFROM:+0200
TZOFFSETTO:+0200
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251026T030000
RDATE:20261025T030000
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20260329T030000
RDATE:20270328T030000
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Your App Just Got a Second User — and It Doesn't Need Its Own UI
  - Sanjay Krishna Anbalagan
DTSTART;TZID=Europe/Luxembourg:20261007T173000
DTEND;TZID=Europe/Luxembourg:20261007T180000
DTSTAMP:20260916T164551Z
UID:pretalx-open-source-conference-luxembourg-2026-ZZV37G@pretalx.com
DESCRIPTION:Your application has a second user now: an AI agent sharing th
 e same session as the human. It can navigate\, select\, fill\, and act\, b
 ut it should not need a second interface — or a second permission model 
 — to do it.\n\nMany agentic applications solve this by generating new in
 terfaces or by exposing application functions directly to a model. Both ap
 proaches discard something valuable: the semantics\, accessibility\, valid
 ation\, permissions\, and interaction patterns already encoded in the appl
 ication.\n\nThis talk presents a different model. Chat becomes an intent s
 urface\, while results and human intervention remain in the application's 
 existing UI. Controls declare what they do\, when they are available\, wha
 t state they change\, and whether the decision belongs to the human or the
  agent. The agent operates through existing application handlers\, while t
 he application remains authoritative about permissions and state.\n\nThis 
 becomes especially important during multi-step agent execution. A button o
 ffered to an agent five seconds ago may no longer be valid. A human may co
 mplete part of a workflow while the agent is waiting. Sensitive informatio
 n can remain entirely inside the application's existing controls instead o
 f traveling through the model context.\n\nWe will walk through these inter
 action patterns and the runtime evidence needed to distinguish “the agen
 t said it worked” from “the application actually changed.”\n\nThe go
 al is a model-independent architecture in which the AI interprets intent a
 nd plans actions\, while the application continues to own its UI\, state\,
  permissions\, and truth.
LOCATION:Artificial Intelligence
URL:https://pretalx.com/open-source-conference-luxembourg-2026/talk/ZZV37G
 /
END:VEVENT
END:VCALENDAR