Oracle's Cloud Prediction #7 for 2020: Data scientists will be in progressively higher demand but will become more efficient due to AI-driven data gathering and analytics techniques.
Each day the world creates 2.5 quintillion bytes of data, and 90 percent of the data in the world was created in the last few years, according to Forbes. This growing treasure trove of knowledge can bring transformative benefits to the world and businesses—if we can keep up.
We already have an established discipline for this (data science) and a new class of professionals to do it (data scientists). But because data science is a relatively new field and the demand for data scientists far exceeds supply, organizations might not be able to use all of their data to its fullest potential. So how can they fill the gap within the next five to 10 years and make the most of the data scientists they do have?
Technology is the answer, and Oracle has a solution for this called the Oracle Data Science Platform specifically.
Subscribe to the Oracle Analytics Advantage blog and get the latest posts sent to your inbox
Data scientists are curious, analytical, methodical scientists who make discoveries with data. They use scientific methods, processes, algorithms, and systems to extract value from disparate sets of data. They have a wide range of skills, including statistics, computer science, and business knowledge. They analyze, reveal trends and—here's the big pay-off—advise on business decisions from data collected from internal and customer-facing sources, as well as third-party sources.
Data scientists work in every industry, and their value comes from knowing how to use machine learning (ML) algorithms to find answers and insights that align to business use cases. They have a different skill set than computer engineers, analysts, data engineers, and ML engineers—but they could work on teams that include any or all of these other roles. Advanced technology has made it possible for teams of data professionals to provide businesses, research groups, government agencies, and other organizations with more accurate and comprehensive intelligence than in the past.
Data scientists can conduct descriptive, predictive, and prescriptive analytics. Descriptive analytics uses current data to describe the current status of a variable and apply that knowledge to future models for improved results. Predictive analytics goes a step further, allowing companies to predict tactical scenarios like a buyers' next move on the purchase journey, as well as planning scenarios like how much production capacity will be needed next year.
ML, a form of artificial intelligence (AI) that identifies patterns in data and learns from them, is a foundational component of predictive and prescriptive analytics.
Prescriptive analytics is an area of high growth. It is defined by my colleague Dr. Elena Drozd as "the missing piece between data scientists and business leaders: The concept of what action you should take right now when predictive intelligence tells you the most likely outcome in the future."
Prescriptive analytics answer these questions:
Demand for prescriptive analytics is quickly escalating because it provides a big competitive advantage. However, a lot of this work is still done manually, and the time constraint limits how much actual analysis a data scientist or data science team can do.
Data scientists typically spend 80% of their time collecting, cleaning, organizing, and preparing data, and only about 20% of their time looking for patterns and discovering new insights. This isn't a good use of their time, particularly when the technology exists to automate and augment analytics.
As AI and ML technologies become more sophisticated, they are automating a larger slice of manual data science tasks. Likewise, as augmented analytics systems (analytics + AI) grow powerful enough to train and execute algorithms at scale, the vast majority of data collection and preparation tasks will be automated, making each skilled data worker more efficient.
AI systems also will get better at generating insights and interpreting results, which will free up data scientists to determine which findings are most relevant out of all the potential outcomes.
Oracle Analytics Cloud is leading in this area by providing self-service access to data and analytics for any role. With built-in advanced AI and ML, the platform as a service (PaaS) automates and eliminates key tasks that would previously have been carried out by IT managers This reduces costs and improves data insights in support of strategic goals.
When Oracle Analytics runs on the Autonomous Data Warehouse, it provides elastic scaling so that companies can expand their analytics services while controlling cost and eliminating the need for specialized database maintenance skills.
Together Oracle Analytics and the Autonomous Data Warehouse can greatly increase data scientists' productivity. They can access their data output from the ML models as well as all available data from across the company. Additionally, they can quickly see visualizations of their data and with embedded AI and ML learn where to focus their future analytic efforts This shrinks the time it takes to derive value from growing volumes of data.
This prediction is one in a series explaining how today's enterprise clouds are evolving—and how they will most likely look in 2025. To read the other predictions, visit https://www.oracle.com/2020-cloud-predictions.
To learn how you can benefit from the Oracle Analytics Cloud, visit Oracle.com/analytics.