Standalone ML vs. ML embedded in Business Applications – Which is better?

Machine Learning (ML) is the buzz words in Business Tech. Every Business is looking to use Artificial Intelligence or Machine Learning to improve its business outlook, operational efficiency, and business process. Every CIO and IT team has the charter to spend on projects to use these emerging technologies as mandated by management, peer pressure, and to stay ahead of the game.

But one of the dilemmas that they are faced is – where and how should they invest? What would be the best use case? Should it be in an ML platform or should it be attempted in the context of the business applications that their users are already familiar with?

As we have evolved in the Business technology space, I always look to history for learnings. Back in the early days when Business Intelligence (BI) became the hot trend, initially, everyone invested in standalone BI platforms. These Data Warehouse and BI platforms worked standalone and continue to exist, but businesses have moved towards getting Analytics in the context of the business process.

I see the same approach for ML. Many businesses have invested heavily in ML platforms, where they have to bring all the data into another store and then run through algorithms, provide the results. For this, the users will go to another system, visualize the data and finally get to another business application to take action using the results – either for business outlook, improving operational efficiency, or reimagining the business process. The major drawback of this approach is the context is lost and the IT teams have to spend an enormous amount of time and effort to bring data science, ML results, and business process altogether.

So what would be a better option – having ML embedded in Business applications, like how BI embedded in Business applications gave great value. This is a better option as the use cases are built in the context of the application e.g. whether it is to get better cash management, or improve sales forecast or initiate preventive maintenance. Also, there are tons of data generated in your business applications – manually entered, or captured through IoT, or processed through different actions. ?

?With our new IPM Insights and other ML investments we are making in Oracle Cloud Applications, we are looking to get ML embedded into our Applications.?This, we see providing the best value for businesses, as it's in the context of the business processes. Context is King.