Algorithmia Survey: Large Enterprises Have Taken the Lead in Machine Learning
Algorithmia announces the results of a survey on enterprise machine learning. The comprehensive survey, titled “State of Enterprise Machine Learning,” is a first for Algorithmia and was designed to explore the ways in which companies of all sizes are utilizing machine learning. The survey was completed by over 500 data science and machine learning professionals, the majority of whom were based in North America. A report detailing the survey’s findings can be found here.
A key takeaway from the survey was that data science and machine learning professionals within larger organizations (2,500+ employees) are feeling significantly more satisfied with their progress than those in smaller organizations—they are roughly 300% more likely to consider their model deployment “sophisticated” and 80% more likely to be “satisfied” or “very satisfied” with their progress as compared to professionals in companies of 500 employees or less.
Following are other key findings from the survey related to machine learning in the enterprise:
- 92% of respondents in companies with 10,000 employees or more said their organization’s investment in machine learning had grown by at least 25% in the past year. By comparison, 80% of respondents in companies with fewer than 10,000 employees said their organization’s investment grew at least 25% in the past 12 months.
- Among enterprises with 10,000+ employees, the biggest machine learning use-case was increasing customer loyalty (59%), followed by increasing customer satisfaction (51%) and interacting with customers (48%). Larger organizations more heavily emphasized the use of data science to save money, too. 48% of respondents in companies with 10,000+ employees cited cost savings as a primary ML use-case, compared to 43% in companies of 1,001-2,500 employees and 41% in companies with 2,501-10,000 employees.
- Large tech companies have a head start over their competition when it comes to machine learning. These companies have created a new category of infrastructure–which Algorithmia refers to as the “AI Layer”– to manage compute loads, automate the deployment of machine learning models, and provide tools for managing machine learning across the organization. Some examples of AI Layers created by big tech include FBLearner from Facebook, TFX from Google and Michelangelo from Uber.
“In 2018, large enterprise companies have an advantage when it comes to machine learning because they have access to more data, can continue to invest in big R&D efforts, and have many problems that machine learning technology can solve cost-effectively,” says Diego Oppenheimer, CEO at Algorithmia. “And yet, even in the largest companies, productionizing and managing machine learning models remains a challenge. Productionizing models is seen as the last step to ROI. Without an enterprise platform to help, these companies are missing out on the rewards of machine learning.”
Resource and Infrastructure Issues
Despite the fact that machine learning is benefiting from massive investments of time, money and focus, a variety of production challenges remain. For example, data science and machine learning teams are spending too much time on infrastructure, deployment and engineering, and not nearly enough (less than 25%) on training and iterating models. Other challenges include:
- 38% of respondents reported difficulty in deploying models to the needed scale. Anecdotally, the reasons include: DevOps and IT teams not having sufficient resources; data scientists being expected to build the infrastructure to put their models into production; and a lack of existing infrastructure within the organization to support the needs of running ML models at scale.
- 30% of respondents reported challenges in supporting different programming languages and training frameworks. Machine learning models often are created using a number of different programming languages and training frameworks. This adds an additional level of complexity because several models, written in distinct languages and frameworks, must be pipelined together for a given task. Larger companies, like Facebook and Uber, have solved some of these issues by building internal tools such as FBLearner and Michelangelo, respectively.
- 30% reported challenges in model management tasks such as versioning and reproducibility.
“In general, larger companies have more machine learning use-cases in production than smaller companies,” says Oppenheimer. “But across the board, all companies are getting smarter about where and how to apply ML technology. We expect to see big leaps in productionized machine learning over 2019 as data scientists can more easily deploy and manage their models.”