ExperTalk is a community of Sinequa Experts (Partners, Customers, Sinequa, Thought leaders) fostering technical exchanges among members to enhance know-how and good practices.

From detecting confidential content among data silos, to processing customers requests or categorize adverse event reports, a world of new opportunities and use cases can be addressed with Machine Learning Classification algorithms. In this talk, we will walk you through the methodology and Sinequa tools required to industrialize the usage machine learning algorithms in production applications. Among other, you will learn:

June 18th, 4:00PM CET / 10:00AM EST*

During this session, Adrien Gabeur, Director of AI Solutions at Sinequa will explain how Sinequa relabeling application lets selected user groups or data experts securely navigate content with Sinequa cognitive search capabilities to perform the following key tasks:

  • When no training set is available to train a model, collaboratively relabel documents to build training sets from scratch (commonly referred as a cold start)
  • When a training set already exists, continuously enhance the training sets, to ensure it stays up to date and prevent data drift.
  • Analyse models predictions and predictions’ confidence and build validation set, to further measure control and monitor the business output of models.
How to automate your machine learning production life cycle, that is:
  • Automatically re-train your models, to leverage new training data generated through relabeling applications
  • Automatically validate your models before deploying to production
  • Swap models instance to deploy latest version online and apply in bach or real time

We will also answer your questions and take your feedback.

Share with your peers!

Sinequa Partner Corner Sinequa University

Register now

Share this