DeepTech Dialogue: Defining the MLOps Playbook

Businesses are embracing digital like never before with the adoption of new-age technologies like AI/ML, cloud, IoT, etc. Especially though MLOPs, they are combining the potential of Machine Learning, DevOps, and Data Engineering to deploy and maintain ML systems in production. As a result, these businesses are able to increase revenues by reducing the time taken in delivering ML products to the market, improving their ROI on AI/ML & Analytics Initiatives, etc.

However, current Machine Learning processes are still considered to be complex in nature with challenges around its reproducibility, auditing, and compliance, validations of the model output, and monitoring for the integrated systems.

Research also highlights how while ML models are growing exponentially, only 27% of the pilot projects are into production. Hence, there is a need to address the challenges/hurdles that are limiting businesses whilst leveraging ML and unlocking its true potential.

In this episode of DeepTech Dialogue, esteemed panelists:

-Gope G Biswas, Asst. VP-AI/ML Implementation & Deployment Lead, Genpact

-Swati Jain, VP, Analytics, EXL

-Sanjay Kukreja, Global Head of Technology, eClerx and Advisor, NASSCOM

throw light on the benefits of MLOps, the implementation pillars and need for control and governance for MLOps processes. They also discuss the industry use cases, strategy enterprises should adopt to move the pilot projects into production and the future of MLOps.