Navigate data management challenges to enable AI initiatives | Strategy, Analytics & M&A

Article

Navigate data management challenges to enable AI initiatives

Smart data management is the foundation of organisation-wide usage of Artificial Intelligence

Leading organisations are able to fully leverage the power of Artificial Intelligence and generate value by enabling data professionals to have access to well-organised high quality data from across the entire organisation. But how can this be achieved?

The Deloitte AI Loop (DAIL)

The Deloitte AI Loop provides a framework that mimics the human approach in the space of artificial intelligence. Based on our experience in bringing cognitive solutions to our clients, we have lined out DAIL as a blueprint for all aspects that should be covered in a successful AI solution, as we explained in the introductory blog.

This is the second article of the DAIL series, focusing on the SENSE component, consisting of tools, technology and infrastructure to measure, capture and monitor data from business processes, behavior and the environment.
 

Deloitte AI Loop (DAIL) SENSE

Value of SENSE

In Deloitte’s latest State of AI in the Enterprise survey of more than 2700 executives, respondents indicated that the modernization of their data infrastructure is their top initiative in order to increase competitive advantage from AI.

This is aligned with our experience showing that key barriers to adoption of AI-driven initiatives originate from challenges in data infrastructure, governance and management. Common challenges are:

  1. Data is spread across different business systems across organization. Organizations are adopting specialized systems to manage an increasing data volume, but these are often tailored to specific business requirements or initiatives within a single line of business, not aligned to the enterprise-wide data architecture. A scattered data environment limits visibility and usability of data assets for data scientists.
  2. Data scientists spend the majority of their time in an AI project to address two factors from the dataset to drive quality of the prediction models:
    • Data quality: high quality data is necessary, otherwise machine learning algorithms will mistakenly learn to mimic and then produce inconsistent data results. This results in weak performance at the least and harmful outputs at the worst. Data duplication, incompleteness and inconsistency decrease the value of the resulting AI applications and increase the amount of time data scientists spend in trying to fix it.
    • Data context: data capturing the context of the problem is as important as data capturing the outcome. Data scientists often work very closely with SMEs (subject matter experts) to make sure the contextual data is also provided to the model. For example, for a manufacturing use case, to perform automated quality control, it is not sufficient to only know the final quality label. Relevant supporting evidence, such as color, weight, dimensions, hardness, and other testing parameters, also need to be provided to the data, which normally comes from the manufacturing engineers or quality inspectors.

 

To achieve higher data quality and more data context, it’s recommended that the organization invests the time and effort upfront in establishing uniform data foundation with integrated capabilities of data security, data management and data governance. A cloud-based data lake provides a central platform to bring together siloed data sources, to enable AI models to SENSE the environment where the organization operates.

The next section describes Deloitte’s approach for enabling SENSE.

Enabling SENSE

Given the challenges, what are the best approaches companies can take to accelerate preparation of high quality data for analytics and fast-track deployment of AI projects into production?

The figure below shows 4 guiding principles to enable SENSE in organizations. For detailed descriptions please read the full article in the deep dive provided at the top of this page.

While enterprises can still get started with AI without the capabilities listed in the figure, having these practices in place ensures a solid foundation for scalability and organization-wide adoption. In addition, AI models developed on data of higher quality typically deliver deeper insights and more sustainable results, ultimately generating value for the organization.

What to expect next?

This blog is part of a series in which we deep dive in the different components of DAIL and describe them in a more in-depth fashion. Next up, we will discuss the REASON which will bring us from the data we have collected to the models we employ.
 

Did you find this useful?