MLOps Systems at Scale with Krishna Gade

 Although we like to think about ML workflows as straight-line narratives from experiment to training to production, and then finally monitoring; the reality for large companies is that all the steps are happening at one time in concert with other models, with shifting data, and, sometimes, misaligned key feature inputs.

Moreover, regulated firms are required to track all the models, the changes, and the impacts of those changes for compliance. Enter explainability supported by model monitoring. Far from the old, single task process of only monitoring changes and anomalies, today’s ML monitoring does much more. It uses AI to deliver full featured performance management that can identify changes, alert the right people, and pop models back into production in real time with proper governance.

FiddlerAI is a startup focused on enterprise model performance management. They are tackling the unique challenges of building in-house stable and secure MLOps systems at scale. Today we are interviewing Krishna Gade about trusting AI, the technical challenges of ML monitoring and the real world problem statements beyond compliance that explainability can address.

Sponsors

This portion of Software Engineering Daily is brought to you by Azul.

Trusted by engineers and high performing development teams around the world as the only company focused solely on Java. Azul gives you more performance, scale, throughput, and security. More of the Java you love, at the price you can afford. Using an optimized Java Virtual Machine (JVM), Azul Platform Prime turbocharges the performance and scalability of your Java applications with a hyper-optimized runtime that drives down infrastructure costs. Say goodbye to latency lag: download Azul Platform Prime for free.

 

For many of us, driving has become a mundane experience: a chore to get us from point A to B. But it doesn’t have to be boring! Nissan EVs are here to be the change we need.Nissan makes EVs – but not just any electric vehicles. They make EVs that truly electrify. The feeling of being in an electric vehicle is truly a step up from traditional gas vehicles. The smoothness of the ride and the quiet of the engine makes a car ride go from something average to something to look forward to. Nissan EVs are here to elevate your experience behind the wheel.And Nissan knows what they’re doing. Since 2010, Nissan LEAFs have traveled over 8 billion miles globally – that’s the equivalent of driving to Pluto and back. One of their EVs even trekked to the North Pole.Elevate your experience today. Nissan, EVs that electrify. Visit NissanUSA.com to learn more.

This segment of Software Engineering Daily is brought to you by Privacy Dynamics, the simple solution for anonymizing data intelligently.

Development and testing environments are the number one source of costly data breaches. Privacy Dynamics creates a de-identified copy of your production data in minutes, allowing developers to maintain their velocity and improve testing without taking on unnecessary compliance and security risks.

Using a proprietary approach based on the latest research, Privacy Dynamics automatically detects and eliminates PII while maintaining the schema, format, scale, and analytical utility of the source data. With a simple integration into any data store, Privacy Dynamics anonymizes data in Postgres, MySQL, Snowflake, Big Query, S3 and more.

Are you letting sensitive PII land in development and testing environments? Eliminate the #1 threat of a data breach and start using de-identified data in your development and testing environments by signing up for a free account today at http://privacydynamics.io/sedaily.

Software Daily

Software Daily

 
Subscribe to Software Daily, a curated newsletter featuring the best and newest from the software engineering community.