BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//pyconde-pydata-2026//speaker//STRRSM
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-pyconde-pydata-2026-MMQ8SX@pretalx.com
DTSTART;TZID=CET:20260414T114500
DTEND;TZID=CET:20260414T121500
DESCRIPTION:RAG-based AI agents fail in production because retrieval withou
 t memory is like a conversation with someone who forgets everything you've
  said. This talk introduces a memory architecture that transforms how you 
 build AI applications with a Python SDK.\n\nUsing an open-source Python SD
 K\, cognee\, I'll demonstrate how to replace fragile RAG pipelines with a 
 unified memory layer combining knowledge graphs and vector search. You'll 
 see live code showing how 6 lines of Python can give your agents persisten
 t\, queryable memory that survives restarts learns and improves with inter
 actions.\n\nWe'll build a working agent memory system using cognee\, Kuzu\
 , LanceDB\, and your choice of LLM provider. The graph and vector layers r
 un embedded with zero infrastructure setup\, no database servers required.
  By the end\, you'll understand why the future of AI agents isn't better R
 AG but better memory.
DTSTAMP:20260412T141733Z
LOCATION:Helium [3rd Floor]
SUMMARY:AI Memory: From Stateless RAG to Persistent Knowledge Graphs in 6 L
 ines of Python - Hande Kafkas
URL:https://pretalx.com/pyconde-pydata-2026/talk/MMQ8SX/
END:VEVENT
END:VCALENDAR