BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.pydata.org//pydataglobal2025//speaker//JWQCRC
BEGIN:VEVENT
SUMMARY:Streaming AI Workflows in Python: Kafka Queues and Flink-Powered L
 LM Inference - Shekhar Prasad Rajak\, bhrathjatoth
DTSTART:20251211T133000Z
DTEND:20251211T140000Z
DTSTAMP:20261006T024259Z
UID:pretalx-pydataglobal2025-7MEX7V@cfp.pydata.org
DESCRIPTION:Python users working on real-time analytics—from payment pro
 cessing and fraud detection to AI-driven support—rely on message queues 
 to keep data moving reliably and efficiently. Traditional message queues\,
  however\, can struggle with large-scale\, concurrent workloads\, especial
 ly when you need durability and replayability.\n\nIn this session\, we’l
 l show how Kafka 4.0 introduces robust queue semantics to distributed stre
 aming\, empowering Python applications to handle fair\, concurrent\, and i
 solated message processing at scale—using familiar Kafka Python clients 
 and frameworks.\n\nBut the power lies in what you can build next. We’ll 
 demonstrate how Apache Flink can connect Kafka event streams to real-time 
 Large Language Model (LLM) inference for tasks like sentiment analysis and
  summarization\, all orchestrated via Python APIs and remote model endpoin
 ts for powerful\, flexible AI inference.\n\nTo complete the picture\, we
 ’ll cover how enriched results can be stored in popular data lake soluti
 ons—such as Apache Iceberg—enabling long-term analytics\, time travel
 \, and integration with downstream data science workflows. Support for Ice
 berg and other lakehouse formats is optional\, giving you flexibility to c
 hoose the right data backend for your needs.
LOCATION:Machine Learning & AI
URL:https://cfp.pydata.org/pydataglobal2025/talk/7MEX7V/
END:VEVENT
END:VCALENDAR