Often the most difficult and costly step before promising but complex enterprise technologies becomes mainstream is for organizations to figure out how to operationalize them. Knowledge graphs and machine learning are both at this stage. How do they fit in with existing software, systems, processes, and resources? What has to change, what will it cost, how long will it take?
Big questions, but both articles this week should be useful when you’re thinking about strategies. First, Kabir Nagrecha discusses how existing systems research and practices contribute to machine learning operations (MLOps). Next, Paco Nathan looks at ways to think about knowledge graphs in the context of systems thinking and complex problem solving.