Q3 Ardoq AI Roundup: Elevating From Productivity to Governance

In Q2, Ardoq embedded AI directly into core modeling tasks to accelerate productivity. But speed in outputs alone wasn’t enough. By Q3, the challenges we were hearing from Enterprise Architects had shifted:

• Exploding scale: data, systems, models, and AI Agents were growing too fast to keep up with manual documentation methods.
• Underused insights: reports and dashboards were becoming hard to navigate for business stakeholders, especially if they had a dense library of them. 
• Shadow AI: AI tools and agents were popping up across the enterprise with no visibility or governance, resulting in exposure to risk. 

Thus, our Q3 AI launches focused on making insights easier to access, giving architects and stakeholders new ways to discover inefficiencies, and creating a governance layer so enterprises can adopt Agentic AI with confidence.

Here’s a look at what we launched, why it matters, and how it all connects.

Announcing AI Lens (Enterprise AI Governance)

The First EA Solution for Agentic AI Governance and Visibility

Currently in development, available early Q4 2025. To gain access, contact your CSM.

AI adoption is exploding inside enterprises. But most organizations can’t answer basic questions like:

• Where are AI Agents being used in our estate?
• Who owns each AI-enabled system?
• Is this compliant with the EU AI Act or internal data privacy rules?

That’s why Ardoq announced the Enterprise AI Governance Solution: AI Lens.  It’s the first EA solution that connects Agentic AI usage to ownership, compliance, and business value.

With AI Lens, you can:

• Build an inventory of AI-enabled applications.
• Map AI to capabilities and processes.
• Track ownership, lifecycle, and risk.
• Assess compliance against emerging regulations.
• Visualize AI adoption with dashboards and radars.

Why it matters for EAs: This isn’t just an IT governance issue. AI is changing business operating models, and architects are uniquely positioned to show how it connects to value, risk, and transformation.

👉 Read: Ardoq Launches Enterprise AI Management Solution to Help Organizations Govern and De-Risk AI Adoption
👉 Watch: Ardoq Illuminaite: Agentic AI is Everywhere - But Who's Governing the Risk?

ShiftX Introduces Conversational Workflows

Chat with Your Process Map

Available For All ShiftX Customers Now!

Processes are the connective tissue of the enterprise, but traditional process maps are often static, dense, and difficult to align with EA data. Architects and process analysts spend hours manually clicking through diagrams, hunting for bottlenecks or compliance checkpoints.

With the new AI-powered ShiftX natural language interface, you can now build and interact with process maps in plain language.

Why it matters for EAs:

• Speed & accessibility: Non-technical users can ask questions directly, making processes understandable to those outside the process team.
• Better insights: AI surfaces hidden inefficiencies, dependencies, and risks that static diagrams conceal.
• Reduced rework: Adapt workflows with natural language instructions instead of manual edits.
• Cross-team alignment: Business stakeholders can finally interact with processes in a way that feels intuitive.

👉 Read: Conversational AI in Ardoq ShiftX: EA & Process in Collaboration
👉 Learn More about ShiftX Process Modeling in Ardoq

Announcing Chat With Ardoq

Ask In Plain English, Get Permission-Aware, Contextual Insights

Available in Q4 2025. To gain access, contact your CSM.

Even the best reports and dashboards can be dense. We are proud to announce our Q4 plans for launching a Natural Language Assistant in Ardoq. This makes it easier for architects and business stakeholders to interact with architecture data.

Ask questions like:

• “Which systems support our sales onboarding process?”
• “Show me the initiatives linked to the Customer Experience capability.”
• “Summarize this capability heatmap in plain language.”

Under the hood, Ardoq’s assistant will be schema-aware. It will understand your metamodel (Applications, Capabilities, Processes, Initiatives, etc.) and retrieve context accordingly. This means more accurate answers. Because it’s built natively into Ardoq, it also respects your existing permissions and masking rules.

Why it matters for EAs: Unlike generic copilots, Chat with Ardoq is tuned to your enterprise graph. That means fewer hallucinations, clearer answers, and insights that are explainable and auditable.

AI Gateway (MCP Server) Goes GA

From Simple Retrieval to Insight and Blindspot Discovery 

Available Now!

Earlier this year, we introduced the Model Context Protocol (MCP) in beta, making Ardoq the first EA platform to support this open standard. In Q3, we took MCP to General Availability. 

With the GA of our AI Gateway (MCP Server), we’ve launched additional capabilities, such as: 

• Query not just components, but full viewpoints, dashboards, and even the metamodel.
• Move from simple retrieval (“What’s the owner of this app?”) to discovery questions (“Where do we have overlapping applications across business units?”).
• Generate ad-hoc visualizations from natural language queries.
• Use the assistant of your choice: Claude, ChatGPT, Gemini, or Microsoft Copilot.

Why it matters for EAs: MCP makes your enterprise model accessible where your stakeholders already work. Instead of exporting slides or running workshops, you can provide leaders with a way to self-serve insights in Teams, Slack, or their preferred AI tool, all while keeping the architecture as the source of truth.

👉Read More: Learn How To Use AI Gateway (MCP Server)
👉Watch Demo: AI Gateway (MCP Server) First to Market in Enterprise Architecture

AI Visual Importer: From Picture to Data In Seconds

Turn Any Diagram, Slide, or Sketch Into Living Architecture

Available in Q4 2025. To gain access, contact your CSM.

Architects don’t start with a clean slate. They inherit all kinds of visuals: Lucidchart diagrams, Visio flows, PowerPoint slides, whiteboard sketches, and even photos of sticky notes. This can provide valuable context, but it’s locked in static images that can’t be queried, linked, or kept up to date. This means rebuilding them into a structured model takes days.

With the new AI Visual Importer technical preview, you can drop in a diagram, slide, or even a photo of a napkin sketch, and AI instantly turns it into structured data. The Visual Importer:

• Detects components and relationships from almost any visual format.
• Creates a living model in seconds, mapped to your metamodel.
• Preserves the logic of the original sketch while making it queryable and traceable.
• Embeds the result directly into dashboards, reports, or value streams.

Why it matters for EAs:

• Accelerates onboarding: Transform legacy diagrams, slides, or workshop outputs without re-entry.
• Preserves institutional knowledge: Capture insights from informal sketches or presentations before they’re lost.
• Improves collaboration: Business teams can work in the tools they like, while architects bring that work into the enterprise graph.
• Scales faster: Shifts effort from transcription to analysis and design.

👉Read about our Lucidchart AI Importer (a similar feature)

AI Report & Viewpoint Descriptions

Plain-Language Summaries for Instant Clarity

Available Now!

Reports and viewpoints are powerful, but they can be dense and hard to scan, especially for stakeholders outside the EA team. With AI-generated descriptions, every report and viewpoint now comes with a plain-language automatic summary.

That means:

• Faster onboarding: Stakeholders instantly understand what a report shows.
• Smarter AI integration: Descriptions make it easier for MCP and other AI assistants to retrieve and surface the right insights.

Why it matters for EAs: No more underused reports gathering dust. By making outputs self-explanatory, AI ensures your architecture insights are easier to share, easier to find, and easier to act on.

👉 Read: AI-Generated Report & Viewpoint Descriptions

AI Solution Augmentations Go GA

From Blank Canvas to Structured Model in Minutes

Available Now!

First introduced in open beta last quarter, AI augmentations are now generally available.

With just a few clicks, architects can generate a tailored value proposition, draft end-to-end value streams, create multi-level capability maps, and link them to supporting capabilities — all within a governed workflow that keeps humans in control.

Why it matters for EAs: Instead of starting from scratch, architects can move straight to analysis and alignment, saving weeks of manual work.

👉 Demo: How Ardoq Uses AI to Generate Business Value Streams in Minutes
👉Solution: Business Value Streams

Explore What’s Next in Ardoq Labs

We’ve been experimenting with what’s coming next in Ardoq Labs. Labs is where we share early concepts, exploratory AI features, and our evolving AI roadmap. Think of it as your window into the future of Enterprise Architecture:

• Test-drive experimental capabilities before they’re generally available.
• See how we’re thinking about topics like exploratory roadmaps, AI extensibility, and context-aware agents.
• Share feedback that directly shapes what we prioritize and ship.

👉 Explore: Ardoq Labs
👉 Give Feedback: Ardoq Labs Roadmap

Closing Thoughts

In Q2, our AI developments focused on modeling speed. But Enterprise Architects know AI can’t just be about productivity. It must be accurate, explainable, and governable; otherwise, it won’t scale.

Together, our Q3 launches address each of the EA challenges we identified at the start of this blog:

• Exploding scale → AI Visual Importer and Solutions Augmentations accelerate modeling and absorb legacy diagrams, sketches, and slides so architects can keep pace.
• Underused insights → Chat with Ardoq, AI Report & Viewpoint Descriptions, and AI Gateway (MCP Server) democratize access to architecture insights.
• Shadow AI adoption → AI Lens (Governance Solution) introduces governance, ownership, and compliance so enterprises can embrace Agentic AI with confidence.

Through these launches, AI at Ardoq has evolved to focus on trust, discovery, and governance, moving beyond “wow” moments and toward technology that enterprises can trust, act on, and integrate into the fabric of their decision-making.

Thanks for reading if you've made it this far… and see you next quarter! 

📖 Take it a step further: check out our Prompt Engineering White Paper — the first guide designed specifically for Enterprise Architects to apply Prompt Engineering to their workflows. It’s a market-first resource to help you build confidence and unlock value from AI right now.