Wednesday, September 16, 2026
5:30 PM – 7:30 PM EDT
Self-Correcting ML and Multimodal Search with Elastic
About this event
Two technical talks at Elastic's Ballston office: Kritika Berry on ML pipelines that detect drift and correct themselves, and Josh Phifer on building multimodal search over text, images and video using Elastic and Jina.
The Elastic Washington DC user group hosts two talks at Elastic's own Arlington office in Ballston.
Kritika Berry presents "Dashboards to Decisions: A Self-Correcting ML Loop with Elastic Observability," on automating the monitoring of machine learning pipelines so that logs, metrics and traces feed back into the system to detect drift and trigger correction rather than just raising an alert for a human.
Josh Phifer, Principal Solutions Architect at Elastic, follows with "Building Multimodal Search with Elastic and Jina." It includes two live demos: searching an artwork collection across combined text and image queries with agent-assisted exploration, and finding specific scenes or objects inside video clips using natural language. He covers the implementation tradeoffs behind both.
Schedule is doors and food at 5:30, first talk at 6:00, second at 6:40 and networking from 7:20. Pizza is provided and the event is free. The office is directly across from Ballston-MU Metro, with Colonial Parking and SpotHero nearby.
Organized by
The Washington DC user community for the Elastic search and observability stack, meeting at Elastic's Ballston office for technical talks on search, machine learning pipelines and multimodal retrieval.