Ardoq and AI: Webinar and Platform Highlights for Enhancing EA

In our recent webinar “Enhancing Enterprise Architecture with AI: An Ardoq Perspective”, Jason Baragry PhD, Head of Product Research and Strategy at Ardoq and Ian Stendera, Chief Product Officer, discussed the opportunities AI will bring to EA and the various ways Ardoq is looking at converting ideas into reality.

AI promises to profoundly affect Enterprise Architecture and how companies plan and execute their strategies, and it is already generating significant attention. Approximately 60% of our webinar participants who responded to the opening poll said they are already experimenting with AI or implementing AI solutions in their organizations.

Here are the high-level highlights from our on-demand webinar on AI and upcoming platform updates that will leverage this powerful emerging technology.


Ardoq’s Two-pronged Approach to AI

We have been actively exploring AI integration into the Ardoq platform while working with customers, partners, and developers to understand how businesses can leverage emerging technologies. Different companies have different appetites for AI - some are rapidly building new solutions while others are still waiting to see where there is the best return on investment (ROI).

“While incorporating AI into the Ardoq platform is a high priority, we are not just focusing on AI for the sake of doing it. We are focusing on how AI can actually help companies make decisions and move faster, and then we make sure we can support the techniques necessary to do that.”
- Jason Baragry, Head of Product Research and Strategy at Ardoq

This means looking into two broad areas: 

• How can Ardoq help enterprises experiment with AI and make use of AI? 
• How can AI be used in the Ardoq platform to speed up model creation, improve model analysis, and support decision-making with smarter reasoning?

How Ardoq Can Help Enterprises Leverage AI

Ardoq can help organizations manage their AI innovation and answer questions like:

1. Who is involved in the initial stages of AI ideation?
2. Which technologies are under experimentation?
3. What are the potential benefits, return on investment, and feasibility?
4. How far away is the technology from being able to be used in production?

Read our in-depth, step-by-step guide on How to Manage AI Innovation with Ardoq.

Managing risk and compliance when it comes to AI is also an area of substantial interest. The questions and guide above also indentifies ways the Ardoq Platform can support new and emerging AI Risk Compliance standards like ISO 42001 and the EU AI ACT.

In addition, Ardoq will offer a new out-of-the-box Solution for Application Risk  Management within the next quarter. This will be the basis for organizations seeking to manage AI Risk Compliance.

Jason also introduced possible ways to extract information from Ardoq to support enterprise AI Experimentation and improve the quality of responses from Large Language Models (LLMs).

Read our full article on Retrieval Augmented Generation (RAG) and how to include Ardoq in a RAG pipeline.

Opportunities and Requirements for AI in Enterprise Architecture

In the webinar, Jason and Ian also discussed the various types of AI and provided examples of how EA platforms like Ardoq could apply these techniques to improve strategic business execution.

This included the opportunities and challenges for: 

• Rule-based Learning
• Machine Learning
• Deep Learning
• Generative AI and LLMs

Key takeaways include:

• Automation through rule-based techniques is the most easily applied technology.
• Large language models are rapidly improving on reasoning benchmarks.
• In a B2B setting, AI must produce reliable and accurate results - so it's imperative that critical information isn’t missing from EA models for AI to be an effective aid.
• Deep learning and fine-tuning of LLMs requires homogenized data and learning data sets - something we intend to explore with our customers.
• We are exploring ways of making use of Generative AI to exploit its reasoning capabilities.

Where AI Can Be Applied to EA

Jason and Ian also discussed the challenge EAs have in maintaining IT asset information and business structure information and how this limits architects’ ability to focus on strategic change and design and bringing people along on the journey. They also touched on possible ways to reduce the need for tedious manual work in Enterprise Architecture and augment the EA practice with AI.

Jason adds that in his experience, organizations never have an oversupply of architects. There is always pressure on the EA team’s time. Every architect wants to spend less time on maintaining and governing the architecture and more time on helping the business achieve their strategic objectives.

Some of the ways AI can be used by EAs to reduce manual tasks and increase efficiency include: 

• Extracting unstructured information from workspaces or documents and making it structured and available for modeling in contexts that makes sense for the company
• Automatically finding duplicates in data from imports
• Converting text from natural language (from a non-metamodel way) into a formalized model
• Being able to find owners and assign ownership based on extraction of data and bring together people in communities
• Using deduction capabilities or other techniques to enforce architecture policies, find EA Smells or uncover violations of EA principles
• Automatically suggest priorities or topics which will accelerate strategic execution

“AI is not just about automating processes, it’s about extracting information and being able to use it in a better way in decision-making processes.”

New AI Capabilities in Ardoq

In the final part of the webinar, Ian and Jason discuss what is needed to adopt AI and some of the initial steps Ardoq will take during the next months. 

“These are just a few of the ways AI can be applied. These initial steps set the foundation we can build upon. We want to get them into the hands of customers, learn what is most appreciated and build from there.”
- Ian Stendera, Chief Product Officer at Ardoq

These are a few of the many upcoming AI-enhancements planned in our platform:

1. AI-Powered Description in Surveys: Leverages AI to generate and auto-fill descriptions of your application components. When surveys are pre-filled, application owners are more likely to take ownership and contribute to maintaining the EA repository. This is one of many ways AI can be applied to overcome problems of low data quality and increase data completeness.
   
   
2. AI-Driven Best Practice Assistant: Get actionable advice on leveraging Ardoq’s best practices and use cases instantly with our AI-powered assistant. Ask any question for quick, comprehensive answers based on extensive training from the Ardoq Help Center.
   
   
3. Visual Simplification: - Using a rules-based engine to simplify models in Ardoq and suit different stakeholders. For example, a technical integration model can be collapsed to show high-level views.
   
   
4. Simplify AI Analysis: - Avoid siloed working and help large architecture teams follow design, security and availability principles. Identify anti-patterns or issues, or rule violations.

To take advantage of AI in Ardoq, reach out to your CSM and opt-in to these new features.

If you're new to the Ardoq platform and intrigued to know more, get in touch for a demo.