You have developed a machine learning (ML) model. But are you also using it? Despite common belief, the mere development of an ML model is not enough to realize its value. In “Torch”, our new data science series, I will introduce you to MLOps and how this process helps you bring your ML model into production.
What the benefits of MLOps are and how it differs from DevOps and DataOps.
How you achieve a fully automated MLOps process.
Which technical capabilities your organization needs to make MLOps possible.
Want to see this on a real-life example? In my takeaway, I show you how we brought a demand forecasting model into production.
Hi there, my name is Roozbeh, and I am an engineering manager at Accenture. Originally coming from the world of software, I now focus on the intersection of DevOps and machine learning and on enabling solutions that are reliable, scalable and auditable. Apart from ML engineering, I enjoy hiking and outdoor activities. Thereby, I can disconnect from the digital world and get inspired by nature. PS: Did you enjoy my “Torch” episode? Then let’s get in touch – I would love to pass the torch and discuss the world of data science further!
What I share with you
After development is finished, organizations have to take another step to realize the value of their machine learning model: they have to bring it into production by means of a well-structures MLOps process. For a retail company, we applied this process on a demand forecasting model. What the challenges of this type of use case are and how we specifically dealt with them? You will find all answers to these and more questions in my takeaway.
