Machine learning (ML) is now mission critical in every industry. Business leaders are urging their technical teams to accelerate ML adoption across the enterprise to fuel innovation and long-term growth. But there is a disconnect between business leaders’ expectations for wide-scale ML deployment and the reality of what engineers and data scientists can actually build and deliver on time and at scale.

In a Forrester study launched today and commissioned by Capital One, the majority of business leaders expressed excitement at deploying ML across the enterprise, but data scientist team members said they didn’t yet have all the necessary tools to develop ML solutions at scale. Business leaders would love to leverage ML as a plug-and-play opportunity: “just input data into a black box and valuable learnings emerge.” The engineers who wrangle company data to build ML models know it’s far more complex than that. Data may be unstructured or poor quality, and there are compliance, regulatory, and security parameters to meet.

There is no quick-fix to closing this expectation-reality gap, but the first step is to foster honest dialogue between teams. Then, business leaders can begin to democratize ML across the organization. Democratization means both technical and non-technical teams have access to powerful ML tools and are supported with continuous learning and training. Non-technical teams get user-friendly data visualization tools to improve their business decision-making, while data scientists get access to the robust development platforms and cloud infrastructure they need to efficiently build ML applications. At Capital One, we’ve used these democratization strategies to scale ML across our entire company of more than 50,000 associates.

When everyone has a stake in using ML to help the company succeed, the disconnect between business and technical teams fades. So what can companies do to begin democratizing ML? Here are several best practices to bring the power of ML to everyone in the organization.

The best engineers today aren’t just technical whizzes, but also creative thinkers and vital partners to product specialists and designers. To foster greater collaboration, companies should provide opportunities for tech, product, and design to work together toward shared goals. According to the Forrester study, because ML use can be siloed, focusing on collaboration can be a key cultural component of success. It will also ensure that products are built from a business, human, and technical perspective. 

Leaders should also ask engineers and data scientists what tools they need to be successful to accelerate delivery of ML solutions to the business. According to Forrester, 67% of respondents agree that a lack of easy-to-use tools is slowing down cross-enterprise adoption of ML. These tools should be compatible with an underlying tech infrastructure that supports ML engineering. Don’t make your developers live in a “hurry up and wait” world where they develop a ML model in the sandbox staging area, but then must wait to deploy it because they don’t have the compute and infrastructure to put the model into production. A robust cloud-native multitenant infrastructure that supports ML training environments is critical.

Putting the power of ML into the hands of every employee, whether they’re a marketing associate or business analyst, can turn any company into a data-driven organization. Companies can start by granting employees governed access to data. Then, offer teams no-code/low-code tools to analyze data for business decisioning. It goes without saying these tools should be developed with human-centered design, so they are easy to use. Ideally, a business analyst could upload a data set, apply ML functionality through a clickable interface, and quickly generate actionable outputs.

Many employees are eager to learn more about technology. Leaders should provide teams across the enterprise with many ways to learn new skills. At Capital One, we have found success with multiple technical upskilling programs, including our Tech College that offers courses in seven technology disciplines that align to our business imperatives; our Machine Learning Engineering Program that teaches the skills necessary to jumpstart a career in ML and AI; and the Capital One Developer Academy for recent college graduates with non-computer science degrees preparing for careers in software engineering. In the Forrester study, 64% of respondents agreed that lack of training was slowing the adoption of ML in their organizations. Thankfully, upskilling is something every company can offer by encouraging seasoned associates to mentor younger talent.

Democratizing ML is a powerful way to spread data-driven decision-making throughout the organization. But don’t forget to measure the success of democratization initiatives and continually improve areas that need work. To quantify the success of ML democratization, leaders can analyze which data-driven decisions made through the platforms delivered measurable business results, such as new customers or additional revenue. For example, at Capital One, we have measured the amount of money customers have saved with card fraud defense enabled by our ML innovations around anomaly and change point detection.

The success of any ML democratization program is built on collaborative teamwork and measurable accountability. Business users of ML tools can provide feedback to technical teams on what functionality would help them do their jobs better. Technical teams can share the challenges they face in building future product iterations and ask for training and tools to help them succeed.

When business leaders and technical teams coalesce around a unified, human-centered vision for ML, that ultimately benefits end-customers. A company can translate data-driven learnings into better products and services that delight their customers. Deploying a few best practices to democratize ML across the enterprise will go a long way toward building a future-forward organization that innovates with powerful data insights.

This content was produced by Capital One. It was not written by MIT Technology Review’s editorial staff.