Machine Learning models in production: MLOps

Details
We will be hosting this QuantumBlack Middle East Meetup virtually on Thursday, 9th March 2023, 9:00 AST / GMT+3
The session will focus on Machine Learning models in production: MLOps
***
AGENDA (AST, GMT+3):
9:00 AM – Zoom Link Active, Welcome & Introductions
9:10 AM – 9:30 AM – "Automating ML model retraining" by Anil Choudhary, Principal ML Engineer at Quantumblack, AI by McKinsey Singapore
9:30 AM – 9:40 AM – Q&A
9:40 AM – 10:00 AM – "The Evolution of MLOps: Challenges and Solutions" by Xingshen Qian, Senior Architect, MLOps Iguazio, California
10:00 AM – 10:10 AM – Q&A
10:10 AM - Closing

***
SPEAKERS:
Automating ML model retraining by Anil Choudhary
The universal truth about ML systems is that model performance deteriorates over time. This can happen due to several factors, including data drift, assumptions made in the modelling steps etc. The solution to this is known to be very simple - retrain the model. However, the process itself can become maintenance burden and retraining code can get out of sync with main solution. In this session we will talk about automated model retraining and how to solve for the risks from automation in production environments.
Bio:
Anil works as a Principal Machine Learning Engineer at QuantumBlack, AI by McKinsey. He collaborates with Data Scientists and Data Engineers to solve challenges of operationalizing machine learning for large-scale analytics transformations. He is based in Singapore and looking to connect with the diverse community of analytics practitioners globally.

The Evolution of MLOps: Challenges and Solutions by Xingshen Qian
Deploying a machine learning model from experimentation to production used to be a long and painful process. There are many challenges for a machine learning project, which involves many teams such as data engineering, data science, application development, and DevOps. The teams on the project each have their own processes and use different tools that created a lot of re-work and friction to deploy a mode to production and generate ROI for the business.

With the new MLOps methodology, a machine learning project will be carried out by designing a continuous pipeline to perform many steps in an ML project life cycle. Many MLOps tools have been developed to achieve this goal, MLRun is one of the most powerful MLOps frameworks. MLRun is an open-source framework that streamlines the MLOps process for data engineering, feature engineering, model training, model serving, model monitoring, and CI/CD. We will briefly discuss the advantages of using such a framework for your ML projects.
Bio:
Xingsheng is a Solution Architect with Iguazio and the Community Manager of the Open-Source project MLRun. He works with enterprises on modernizing their MLOps processes and guides clients with best practices. He has deployed many ML projects in a variety of sectors including financial, manufacturing, energy and utilities, transportation, and IIoT. He is based in San Jose, California, and looking to connect with global communities of MLOps.

Sponsors:
This event is sponsored by QuantumBlack, AI by McKinsey.
QuantumBlack is an advanced analytics firm operating at the intersection of strategy, technology and design to improve performance outcomes for organisations.
***

Notice:
By participating in the meetup you agree that McKinsey & Company may (i) videotape, audiotape, photograph, or otherwise record your name, voice, or image, and (ii) use and distribute such videotapes, audiotapes, photographs or recordings of your name, voice or image in any texts, videos, and other materials that McKinsey may make available through websites and social media to its employees or any third parties.
If you do not wish to be recorded please inform the event organiser in advance.