Like new things, we are often drawn to the latest innovations without taking full consideration of the underlying foundations and what is needed to harness these innovations. Recently, “artificial intelligence” and “AI” have become established buzz phrases. It is an umbrella term that includes a variety of technologies and applies to many different use cases. As a result, it is more appealing for many companies to focus on a glorious, yet more limited prototype, rather than investing in the underlying capabilities to enable longevity.
A coordinated approach
There are several important foundations to create useful and sustainable value from everything under the AI umbrella. A coordinated approach to ensuring cohesion across organizational culture, data governance and technology strategies. This is also of the utmost importance, assuming that the work required to assess the need for AI was done in the first place. There are important steps to consider before consistently using AI and machine learning (ML) tools to provide information advantages.
While not a unique issue, the need to lay these foundations is particularly profound in areas of defense and national security, where digital tools are difficult to expand. The path from proof of concept to production-ready tools is financially dangerous, without considering evergreens and support. There are many enablers that organizations can use to scale and maintain AI effectively, but there is definitely something that should be considered essential: a data platform.
Data Platform
Data platforms provide the infrastructure needed to support data applications and turn raw data into strategic assets by leveraging a single source of truth. Trustworthy and reusable data is an important prerequisite for any analysis, especially AI, and by combining it, it must be done in a cyber-secure and access-controlled way.
The term “platform” can be interpreted in a number of ways. Within Leonardo's UK cyberbusiness, the “data platform” is called a digital fabric built with a microservices architecture. This allows for the intake of a variety of data sources, allowing seamless access, integration and scaling. Without a competent, secure data platform, it is impossible to gather data insights, create trustworthy AI systems within digital systems, or control control risks. Data platforms are also a key component in the creation, management and implementation of data governance, one of the biggest challenges faced by many organizations.
In an age where digital SQEPs (well-qualified and experienced personnel) are scarce, automating data management tasks and enhancing platform intelligence and operational efficiency can dramatically reduce the resources needed to perform time-consuming and mediocre tasks such as data cleaning and database identification. As platforms need to evolve with changing operational requirements, the ability to version control of features and track data lineage significantly improves overall pipeline quality and the system's ability to adapt with PACE.
Nameless Hero
Without the right data platform, attempting to analyze big data in any form, particularly using AI/ML, results in a set of concepts that are only suitable for use in isolated environments.
Data platforms are an underrated bedrock that drives the rise of AI. The ability to effectively manage, process and extend data, along with providing the necessary calculations, tools and frameworks, are all critical to delivering successful and secure digital applications.
