How to build a data science and machine learning roadmap in 2022

Data scientist teams in financial services, health care, and manufacturing tell VentureBeat their enterprise’s DSML strategies are the most effective when they anticipate and plan for uneven initial results by business unit. The teams also say producing models at scale using MLOps is fundamentally different from producing mainstream internal apps with DevOps. They add that the more complex the operating model of a business unit, the steeper the MLOps learning curve. DSML’s contributions to business units vary by the availability of reliable data and how clearly defined problem statements are.g

• Adaptive ML shows potential for improving cybersecurity, remote site security, quality management in manufacturing, and fine-tuning industrial robotics systems.