BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.northbaypython.org//ERUJ9T
BEGIN:VTIMEZONE
TZID:PST
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10;UNTIL=20061029T100000Z
TZNAME:PST
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
END:STANDARD
BEGIN:STANDARD
DTSTART:20071104T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:PST
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000402T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4;UNTIL=20060402T110000Z
TZNAME:PDT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20070311T030000
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:PDT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-nbpy-2026-ERUJ9T@pretalx.northbaypython.org
DTSTART;TZID=PST:20260425T154500
DTEND;TZID=PST:20260425T161000
DESCRIPTION:Python powers Netflix's recommendation engine\, processes petab
 ytes at Spotify\, and handles financial calculations worth trillions of do
 llars. Yet\, developers constantly debate whether Python is "too slow." Th
 is paradox reveals a fundamental misunderstanding about Python's true supe
 rpower: orchestration over computation. The real question isn't "Is Python
  fast enough?" It's "What should Python be doing?"\n\nMany developers want
  to use Python when building high-performance systems\, but often lack cle
 ar guidance on where Python is the right choice and where it may become a 
 bottleneck. Without a decision framework\, developers either prematurely r
 ewrite working Python code in C++/Rust (adding complexity without proporti
 onal benefit) or keep everything in Python and hit performance walls\, lea
 ding to systems that are neither fast nor maintainable. This talk cuts thr
 ough the performance mythology to reveal where Python excels and when it s
 truggles. \n\nBy the end of this session\, attendees will leave with a cl
 ear decision framework and practical architectural patterns they can immed
 iately apply to their own projects. Whether you're evaluating an existing 
 "slow" Python service or designing a new system from scratch\, you'll have
  concrete tools to determine where Python belongs and where computation sh
 ould move to compiled languages. You'll know how to structure these bounda
 ries effectively and avoid the common pitfalls that add complexity without
  delivering real performance gains.\n\nKey Takeaways:\n1. Decision framewo
 rk for orchestration vs. computation: Four-question evaluation to determin
 e what belongs in Python versus compiled languages \n2. Three architectur
 al patterns for integrating Python with high-performance code: When and ho
 w to use each pattern effectively (with code examples)\n3. Anti-pattern re
 cognition: Build intuition for justified complexity versus technical debt
DTSTAMP:20260404T081117Z
LOCATION:Barn
SUMMARY:Python as Orchestrator: When to Glue\, When to Compute - Freya Bhus
 han Mehta
URL:https://pretalx.northbaypython.org/nbpy-2026/talk/ERUJ9T/
END:VEVENT
END:VCALENDAR