How are you preparing your Enterprise Architecture for the rise of AI Agents? Agentic systems are redefining automation, interoperability, and governance. Are you integrating agent-driven intelligence into your architecture roadmap or modernization strategy?
Sort by:
AI agents and related technologies are clearly part of the architecture landscape. I see 2 questions here - how are we tracking and managing AI as part of our APM and how are we utilising AI for EA work. I think the intent of the question was probably the former but both aspects are relevant. We use SAP LeanIX for EAM and likely will investigate its Agent hub features. We are also looking at AI use cases for augmenting or automating some types of EA work.
In our architecture roadmap, we have planed Flowable (BOAT/CMT with capabilities on agentic architecture) for the automation and orchestration management. Because we use it for business scenario/process. It has some embedded AI agents that we can use them to help quickly the business.
For the next step, we would like have a LCAP (Low-code application platform) like outsystems or mendix to assist the IA agent development. We wil integrate them with Flowable thru MCP.
We have begun experimenting with building AI Agents for Enterprise Architects - with sub-agents and tools for API generation, diagram creation and research. This is still in the experimental mode, and I would be interested in other peoples view of the applicability of AI Agents for EA work - what do we have to change in how we organize and structure our Architecture data (e.g. adopting diagrams-as-code) to enable AI Agents to be effective.
There is a public reference to our work at https://inform.tmforum.org/features-and-opinion/vodafone-google-cloud-tm-forum-demo-inter-organization-ai-agent-collaboration . In this presentation our internal Agent has the ability to ask questions on Industry standards and specifications by querying an agent developed by Google on the TM Forum standards (the industry standards body for Telco IT).
We have already incorporated Artificial Intelligence as one of our categories for our city wide Technical Architecture Standards which we use for Intake and RFP's. Our AI working group aligned with EA has identified already our SaaS solutions that provide AI agents and we have a roadmap on how to evolve our internal programs and resources on utilizing existing SaaS enterprise solutions that offer AI agents (Salesforce, Workday, Service now to name a few). Our RFP, Intake and VRA process include sections for Artificial Intelligence questions. We are seeing Innovation at the Edge (agencies technical solutions) where individual agencies addressing their business issues are seeing an increase of AI vendors offering AI analysis, AI Bots and AI Services. In one year, the adoption of AI augmentation, Agentic AI is accelerating the next level of innovation in technical architecture. We have several enterprise BOTs in production for several years and as part pf our modernization strategy we did a RFP this year to review AI Services and AI products so we can create a bench of resources. We have had several RFP's focused on AI areas to mature our permit licensing process and intelligent document processing for several agencies. We have been fortunate in that our Mayor and Chief Artificial Intelligence Officer / CIO have demonstrated leadership and vision in the area of AI realizing the benefits for our agencies and constituents. We have embarked on several educational meetings with various vendors to understand their enterprise AI offerings and are using Microsoft CoPilot Chat while running a pilot for CoPilot Office365. Our vendors have been very supportive and collaborative in assisting us on this journey
We are absolutely integrating AI agents into our architecture roadmap. They’re not just a trend but they are reshaping how we think about automation, interoperability, and governance.
Our focus is simple:
-Build a secure control layer so agents operate with clear guardrails.
-Modernize APIs and data access to make them agent-friendly and safe.
-Add observability and evaluation so we can trust what agents do.
-Empower both IT teams and business users with tools to create agents responsibly.
In short, we’re laying the foundation for agents to work alongside people for augmenting decisions, automating workflows, and doing it all with compliance and transparency baked in.