Google, Amazon, Microsoft, and IBM have all jumped into machine learning. What does this mean for your business?

Machine learning (ML) is steadily making its way into enterprise applications across a multitude of industries and transforming the way teams work. In the past two years, major tech companies like Google, Amazon, Microsoft, and IBM have invested millions in building out new machine learning capabilities on their cloud platforms. Their combined investment is radically shaping how ML gets used in organizations and substantially increasing the rate of ML adoption. 

At Encapture, we’ve been diving into this tech to test out capabilities and use cases for our clients. We’ll discuss the results of our benchmarking in a later blog, but let’s first dive into some background and the potential impact on your business.

Why cloud providers are investing in ML

This may sound Machiavellian, but these companies have a single goal in mind: to get more customers onto their cloud infrastructure. All four companies are competing heavily for market share in the $300 billion cloud services industry

Since cloud computing infrastructure is fairly commoditized, their strategy is to build out helpful services and applications within their cloud that are cheap and easy to use. The thinking goes that if they can provide enterprise tools that are closely tied to their infrastructure, it will force customers to become more reliant on their cloud (and create a stickier customer).  

Specifically, these cloud providers have recognized that ML tools are a great fit for their revenue strategy due to the substantial processing power required to perform machine learning. More processing power equals greater cloud infrastructure demand, which results in more revenue. Consequently, in the past several years, all four companies have built out huge teams of data scientists, engineers, and mathematicians to help bring machine learning to the masses.

 The benefits of cloud ML

Thankfully, cloud ML provides several benefits over legacy products.

First, it’s easy to deploy and scale. These ML tools are already embedded in the cloud, and companies don’t have to build out their own infrastructure. Increasing processing power is just a few clicks vs. buying, installing, and managing additional servers.

Second, cloud ML tools are straightforward to configure and tune. Traditional machine learning technology requires advanced skills in AI theory and data science. Cloud tools are designed for business users with a basic understanding of computer language.

Third, the pricing is very simple – you only pay for what you use. This is a nice feature for companies that have uneven or seasonal workloads and need to spike usage at specific time while going idle at other times.

Overall, cloud ML could be a gamechanger for many companies who have struggled to adopt machine learning in the past. It’s opening up new use cases and allowing non-expert users to experiment in an inexpensive environment. And given the amount of continued investment, these products will only get better – we think the rate of change in cloud machine learning over the next couple of years will be greater than all advances over the past decade.

What cloud ML means for your company

Because of cloud ML’s ease of use and low cost, you should be paying close attention to how you can build workflow-specific solutions on this technology, especially if your company already has machine learning from a legacy provider. Given all the options in the market today, you’re no longer tied to a single ML platform and should look to incorporate best-of-breed functionality into your processes.

Importantly, make sure your existing workflow and productivity tools can leverage this technology. A big downside to cloud ML tools is that they function as a service – which means you’ll need “boots on the ground” tech that can manage basic functions like administration, security, content curation, etc. while calling these services. Most capture and workflow vendors DO NOT work well with third-party machine learning tools, as they want you to use their own solutions (or are just too old to integrate).

We’ve built Encapture to work well with all workflow / content systems, including cloud ML tools. Encapture is designed to be a single source of document acquisition and orchestration while integrating with various third-party ML providers to perform data extraction and analysis. While Encapture has its own machine learning that’s easy to use and deploy, we also want to integrate with Google / Amazon / Microsoft / IBM to take advantage of their unique capabilities. We think the world is a better place when systems talk friendly to each other.

Please reach out if you want to learn more about cloud machine learning and how it can be applied to your business! 

Share This