BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.in.pycon.org//2025//3NJXXL
BEGIN:VTIMEZONE
TZID:IST
BEGIN:STANDARD
DTSTART:20000101T000000
RRULE:FREQ=YEARLY;BYMONTH=1
TZNAME:IST
TZOFFSETFROM:+0530
TZOFFSETTO:+0530
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-2025-TK9ULJ@cfp.in.pycon.org
DTSTART;TZID=IST:20250914T153000
DTEND;TZID=IST:20250914T160000
DESCRIPTION:## Can Celery do more than just background tasks? Absolutely.\n
 \nIn this session\, I'll share a real-world use case where we transformed 
 Celery into a full-fledged workflow orchestrator — handling serial and p
 arallel task execution\, task retries\, dynamic routing\, and queue manage
 ment — all running at scale on Kubernetes.\n\nPaired with K8s' elasticit
 y and Celery's simplicity\, this setup replaced a slower\, costlier server
 less architecture built on AWS Lambda and Step Functions. You’ll learn h
 ow Celery’s canvas patterns\, task routing\, and worker isolation let us
  orchestrate complex ML pipelines efficiently\, and why it's still a power
 ful\, underappreciated tool for scalable\, production-grade systems.\n\nIf
  you're tired of serverless trade-offs and want more control without givin
 g up scalability\, this talk is for you.
DTSTAMP:20260317T122428Z
LOCATION:Track 3
SUMMARY:Celery and Kubernetes for a Fast\, Scalable\, and Robust Workflow O
 rchestration - Param Rajani
URL:https://cfp.in.pycon.org/2025/talk/TK9ULJ/
END:VEVENT
END:VCALENDAR