Recent posts by Mona

Trusting in artificial intelligence systems is not easy. Given the variety of edge cases on which machine learning models may fail, as well as the lack of visibility into the processes underlying their predictions and the difficulty of correlating their outputs to downstream business results, it’s no wonder that business leaders often look upon AI with some skepticism. That said, I’ve worked with many forward-thinking AI teams on becoming product-oriented and building trust in their AI systems. I’ve also talked to company leaders about how they can connect their machine learning models to the broader business context in which they’re deployed. At the core of generating trust in AI systems and empowering them to become process-oriented is the concept of model monitoring. However, when monitoring is proposed, data scientists are often confused as to why it is needed, if processes for automatically labeling new data and retraining models are already in place. This kind of thinking is a relic of the traditional, research-oriented approach to data science, and it’s important to understand that the real world is full of scenarios for which model retraining is insufficient.

Data and concept drift are frequently mentioned in the context of ML monitoring, but what exactly are they and how are they detected? Furthermore, given the common misconceptions surrounding them, are data and concept drift things to be avoided at all costs or natural and acceptable consequences of training models in production? Read on to find out. In this article we will provide a granular breakdown of data and concept drift, along with methods for detecting them and best practices for dealing with them when you do.

Mona is happy to announce the expansion of use cases supported by the intelligent monitoring platform for AI / ML to include Robotic Process Automation (“RPA”) workflows. Currently providing customers across 8 different industries with actionable insights into their AI systems, Mona excels at providing complete process visibility, detecting issues within specific segments of data. As a highly extensible platform for many use cases including machine learning, NLU/NLP, speech recognition, and vision, the extension to support intelligent automations and RPA is seamless.  

Monitoring is critical to the success of ML models deployed in production. Because ML models are not static pieces of code but, rather, dynamic predictors which depend on data, hyperparameters, evaluation metrics, and many other variables, it is vital to have insight into the training and deployment process in order to prevent model drift and predictive stasis. That said, not all monitoring solutions are created equal. These are the three must-haves for a machine learning monitoring tool, whether you are deciding to build or buy a solution. 

Before you launch a project to build an AI monitoring system from scratch, consider whether or not this would be a good use of your resources. When does it make sense to buy instead? Let’s discuss. This post explores the advantages and disadvantages of both alternatives so that you can make an informed decision about what’s best for your organization.

We hope that everyone had a fantastic holiday season and is now ready to tackle the 2022 New Year! Looking back to where we started in 2018 to where we are now, we have grown so much overall as a company. From three (⅓ balding) guys with a crazy idea nobody understood, through assembling a team of passionate trailblazers, and to building advanced features for Mona’s platform - now leveraged by incredible AI/ML teams at industry leaders, and even recognized by Gartner, we are continuing to strengthen our position as the leading monitoring solution for AI,  - providing the most flexible and comprehensive insight engine.