How to build an AI-ready organization: the Enterprise Intelligence Architecture

I spoke with IDC’s Marlanna Bozicevich, Research Analyst, Data Platforms, at IDC’s 60th annual Directions conference.  

Marlanna and her analyst colleagues advise senior IT decision makers around technologies such as agentic AI. In our conversation we get the benefit of that advice as we discuss how organizations need to manage data to take advantage of agentic and generative AI.  

Marlanna proposes what she calls the Enterprise Intelligence Architecture, in which data is considered across four planes, laddering up to business activity.  

You can watch our conversation here, or via the YouTube player below. 

AI and data management

I asked Marlanna what is top of mind for the IT leaders with whom she works? 

She said that in the data-management space, there are three general areas where IT buyers are thinking about AI:  

• Productivity: Bringing AI automation into data management to improve operational efficiency (including automating repetitive tasks, and natural language interactions).  

• Responsible AI: Properly managing intelligence about your data.  

• Agentic AI: The next phase of AI maturity, another tool for bringing automation into data management now in the form of autonomous action. In the agentic age real-time data becomes crucial.  

AI for data / data for AI

When we think about data and AI, we tend to focus on the impact AI can have when fueled with quality data. Marlanna is keen that organizations also consider AI as a means of improving data quality.  

I asked her in what ways she sees AI being used to improve data processes and boost productivity among data teams, and how does this impact other parts of the organization?  

She said there are two sides to boosting productivity in data-management practices:  

1. AI for data: Finding ways to improve the processes of data engineers, data scientists, data stewards (adding automated features to improve the data personas workflow).  

2. Data for AI: How the cleaned, AI-ready data can be used internally in other parts of the organization to boost productivity (natural language interface for business personas, data democratization, data access, enhanced collaboration between data and business teams).  

Treat data as a product

That’s managing data within an organization to drive efficiency. I asked Marlanna how can IT buyers use their now efficiently managed data to power AI workloads?  

She said it is important to treat data as a product. Data products are a key success factor for AI initiatives, including agents. In Marlanna’s view data products will be key to facilitating data exchange among agents. Challenges will arise in standardization.  (See also: Can AI solve your technical debt problem?)

Building the AI-ready organization

For any business to generate ROI in AI, inside or out, organizations need to be AI ready.  

Data and the management of and access to data, is critical for the successful implementation of AI projects. I know from my own conversations with IT buyers that there is some tension between what AI can do conceptually, and what AI can do for a living organization, warts and all. IT leaders need to get their organizations AI ready.   

Here Marlanna introduces the concept of Enterprise Intelligence Architecture for the AI fueled business. She says that organizations need to think about data in four planes:   

The four planes of the Enterprise Intelligence Architecture 

1. Data plane. 

2. Data control plane. 

3. Data synthesis plane. 

4. Business activity plane.  

Each plane requires focus and resource so that the organization is capturing, storing and accessing data in a way that is useful to business outputs. Functions across the architecture – the four planes – come together to deliver AI-ready data in the enterprise.  

This structure and approach then supports continuous and collective learning, and the delivery of insights at scale. This in turn promotes data literacy and data culture, supporting continuous development of the Architecture. (See also: How to win at AI: think like a systems designer, not a tech shopper.)