Events

April 29, 2023 | 10:24 am

Meet Elemeta: Metafeature extraction for unstructured data

LLMs are everywhere, left, right, and center of any and all AI discourse these days. But we've got to be honest here, it's hard to understand how they make decisions and explain and monitor them. So earlier this week, we released into beta Elemeta, our open-source library for exploring, monitoring, and extracting features from unstructured data.

View recording

Webinars

March 30, 2023 | 11:25 am

Improving search relevance with ML monitoring

Let's take a dive into ML systems for ranking and search relevance on architectures such as Elasticsearch and vector databases like Pinecone and what it means to monitor them for quality, edge cases, and corrupt data.

Watch recording

Webinars

November 10, 2022 | 4:40 pm

Workshop: A Guide to Putting Together a Continuous ML Stack

We’ll take a hands-on dive into implementing the 1st level of MLOps maturity and performing continuous training of the model by automating our ML pipeline. We’ll start with the ML pipeline and see how we can detect performance degradation and data drift in order to trigger the pipeline and create a new model based on fresh data.

Talks

November 10, 2022 | 4:37 pm

Retraining won’t fix your model (always)

When models misbehave, we often turn to retraining to fix the problem, but retraining is not always the best or only solution out there. In this session we'll take a crash intro in alternative techniques.

Conferences

November 10, 2022 | 3:38 pm

Toronto Machine Learning Summit

Swing by booth #6 and let's talk MLOps and ML observability

Talks

October 11, 2022 | 6:52 am

Data-driven retraining with production insights

In this talk, we'll showcase, through ML monitoring and notebooks, how data scientists and ML engineers can leverage ML monitoring to find the best data and retraining strategy mix to resolve machine learning performance issues. This data-driven, production-first approach enables more thoughtful retraining selections, shorter and leaner retraining cycles, and can be integrated into MLOps CI/CD pipelines for continuous model retraining upon anomaly detection.

Read now

Meet the team and learn about our ML monitoring methodologies and best practices.