BEGIN:VCALENDAR
PRODID:-//AddEvent Inc//AddEvent.com v1.7//EN
VERSION:2.0
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
BEGIN:STANDARD
DTSTART:20261101T010000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20260308T030000
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
DESCRIPTION:Who this session is for\nThis session is designed for Senior Machine Learning Engineers\, Applied AI Researchers\, and Technical Leads responsible for deploying and maintaining production ML systems at scale. It specifically addresses practitioners navigating the transition from research prototypes to systems serving millions of users in high-stakes recommendation\, ranking\, and personalization environments.\n\nWhy you should attend\nIn this talk\, I’ll share practical lessons from designing and operating large-scale recommendation systems in production\, focusing on the decisions and tradeoffs that matter beyond model architecture. Using product substitution as a concrete example\, the session explores how to reason about system design\, evaluation\, monitoring\, and iteration in environments with high data complexity and real-world constraints.\nRather than emphasizing novel algorithms\, this session focuses on production-grade ML thinking: balancing simplicity and complexity\, aligning offline metrics with real-world outcomes\, and building systems that remain reliable over time. Attendees will leave with generalizable best practices and mental models that can be applied across production ML systems in areas such as recommendation\, search\, and ranking.\n\nAhsaas Bajaj is a Senior Machine Learning Engineer at Instacart and recognized expert in large-scale recommendation and retrieval systems. He has pioneered production ML methodologies for high-cardinality decision spaces\, with his work on intelligent substitution and personalization systems publicly recognized in Instacart’s shareholder communications for measurably improving “perfect order fill rate” and customer experience. These systems operate at exceptional scale—processing hundreds of millions of items replacements annually with high satisfaction rates.\n\nPreviously at Samsung Research and Walmart Labs\, Ahsaas has consistently delivered ML platforms under extreme real-world constraints\, establishing methodologies that bridge academic rigor with operational reliability.\n\nAs an invited speaker and thought leader in the production ML community\, Ahsaas focuses on advancing best practices for deploying AI systems that serve millions of users. His work represents a distinctive approach to production-grade ML—emphasizing evaluation architecture\, system design\, and the operational discipline required for real-world impact. He holds graduate training in ML\, NLP and Information Retrieval from UMass Amherst.\n\n------\n\nCreate your own Add to Calendar links with addevent.com/r/a \n
X-ALT-DESC;FMTTYPE=text/html:Who this session is for<br />This session is designed for Senior Machine Learning Engineers, Applied AI Researchers, and Technical Leads responsible for deploying and maintaining production ML systems at scale. It specifically addresses practitioners navigating the transition from research prototypes to systems serving millions of users in high-stakes recommendation, ranking, and personalization environments.<br><br>Why you should attend<br />In this talk, I’ll share practical lessons from designing and operating large-scale recommendation systems in production, focusing on the decisions and tradeoffs that matter beyond model architecture. Using product substitution as a concrete example, the session explores how to reason about system design, evaluation, monitoring, and iteration in environments with high data complexity and real-world constraints.<br />Rather than emphasizing novel algorithms, this session focuses on production-grade ML thinking: balancing simplicity and complexity, aligning offline metrics with real-world outcomes, and building systems that remain reliable over time. Attendees will leave with generalizable best practices and mental models that can be applied across production ML systems in areas such as recommendation, search, and ranking.<br><br>Ahsaas Bajaj is a Senior Machine Learning Engineer at Instacart and recognized expert in large-scale recommendation and retrieval systems. He has pioneered production ML methodologies for high-cardinality decision spaces, with his work on intelligent substitution and personalization systems publicly recognized in Instacart’s shareholder communications for measurably improving “perfect order fill rate” and customer experience. These systems operate at exceptional scale—processing hundreds of millions of items replacements annually with high satisfaction rates.<br><br>Previously at Samsung Research and Walmart Labs, Ahsaas has consistently delivered ML platforms under extreme real-world constraints, establishing methodologies that bridge academic rigor with operational reliability.<br><br>As an invited speaker and thought leader in the production ML community, Ahsaas focuses on advancing best practices for deploying AI systems that serve millions of users. His work represents a distinctive approach to production-grade ML—emphasizing evaluation architecture, system design, and the operational discipline required for real-world impact. He holds graduate training in ML, NLP and Information Retrieval from UMass Amherst.<br /><br />------<br /><br />Create your own Add to Calendar links with addevent.com/r/a <br>
UID:9155faf8fbe64895aca869e9e740bdd5addeventcom
SUMMARY:Ahsaas Bajaj - Production-Grade ML in Practice: Evaluation and Design Frameworks for Recommendation Systems Serving Millions (ML Indsutry)
DTSTART;TZID=America/Los_Angeles:20260130T070000
DTEND;TZID=America/Los_Angeles:20260130T080000
DTSTAMP:20260824T204934Z
TRANSP:OPAQUE
STATUS:CONFIRMED
SEQUENCE:0
LOCATION:https://meet.google.com/fgy-dims-rqj?hs=122&authuser=0
X-MICROSOFT-CDO-BUSYSTATUS:BUSY
BEGIN:VALARM
TRIGGER:-PT30M
ACTION:DISPLAY
DESCRIPTION:Reminder
END:VALARM
END:VEVENT
END:VCALENDAR