How APIs are Opening up the Machine Learning Markets

tech vege/Adobe Stock

Talk of machine learning (ML) has dominated the inner sanctums of heady supply chain discussion of late. While there’s no consensus on where the first game-changing applications are likely to emerge, most everyone understands that running complex algorithms against big data spells growth for the predictive analytics industry. And with the ML providers now offering their products on a subscription basis, well, it’s not hard to imagine where things are likely headed.

If poor data quality and lack of integration have been the ML industry’s proverbial wet towel, then the improvement of APIs and open sourcing seems to be providing more than adequate offsets. ML provider Apigee, whose CTO Anant Jhingran is dedicated to developing and managing better APIs, sheds further light:

“There needs to be a consistent data stream of signals about the behaviors and actions of the end users from all channels of engagement. You can use the data to generate really deep insights with machine learning, then feed the results back into the apps to make improvements.”

Not surprisingly, Amazon, Google, IBM and Microsoft are the biggest players battling to dominate the machine learning cloud services market. APIs that offer capabilities including image tagging, face recognition, speech recognition, predictive analytics and sentiment analysis are the top draws. Other notable APIs include api.ai, Cogito, DataSift, iSpeech, Microsoft Project Oxford, Mozscape and OpenCalai.

Regardless, the procurement and supply chain management marketplace has an opportunity to step forward, as payment and sourcing platforms, even supplier performance management (SPM) apps, are increasingly extra-networked. They support precisely the kinds of important, continuous, data-driving interactions that these ML systems require to be effective. Once a theoretical pursuit, the optimal clearing of electronic marketplaces seems well within reach.

The analysts are convinced that a combination of ML and predictive analytical apps will soon become commonplace line items on most large organizational budgets. Whether that’s true will largely depend on how well coached the evaluators of such systems will be, because the overarching message here isn’t likely to change: users must avoid becoming enamored with the technology and keep their focus on specific objectives and solving specific business problems.

Jhingran provides an even more sobering reminder. “No matter how good the algorithm, no matter how good the scientist, the models can’t perform magic. No data in, no science out,” he joked.

If procurement and SCM professionals are going to provide the goods, they would be wise to ensure their ability to capture the spoils.

Share on Procurious

Discuss this:

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.