Stakion Joins Comet

Over the past 2 years, there has been a steady increase in investment towards Machine Learning initiatives. When we started Stakion, our philosophy was that in order for Machine Learning initiatives to be successful, companies needed to have visibility not only on the expected performance of the models they relied upon, but also the observed performance of these models.

That’s why Stakion has been focused on building a product centered around two concepts: granular monitoring and actionable insights. In order to monitor models effectively, Data Science teams need to have access to a lot of different metrics with varying levels of granularity. In addition, once issues have been identified Data Science teams need tools to allow them to find the root cause of the issue.

That’s also why we’re very excited to announce that Stakion will be joining Comet and bringing ML production monitoring to an ever larger number of companies.

By joining Comet, we are able to add another dimension to production monitoring. By bridging the gap between experimentation and production, teams will now be able to not only identify and find the root cause of issues but also track how the model was adapted to prevent the issue from occurring again. Data Science teams will now have full visibility of the lifecycle ML models in a single platform.

We couldn’t be more excited to start integrating production monitoring into the Comet platform and looking forward to the months ahead. Curious about what’s next for Stakion? Check out the Comet MPM announcement and product page.