Welcome to AI City: Understanding ServiceNow's AI terminology Through a Smart City Story

[sachin45]

What if every AI capability in ServiceNow was a citizen working together in a smart city?

In this article, we'll take a tour through AI City, meet each of its citizens, understand what they do, what they are for, and explore a real-world example of how they contribute to every AI interaction.

🚦 The Gatekeeper | Generative AI Controller

What it does

The city's traffic controller that receives every request and intelligently directs it to the right AI service, ensuring it follows the best route.

What it is (for understanding)

• It understands context and routing rules to direct traffic efficiently.
• Think of it like an airport traffic control system—it knows which runway each flight should use.

Real-world example : A user types a request into Now Assist. The Gatekeeper evaluates it and decides

• "This is a data lookup question → Send it to the Librarian."
• "This needs intelligent reasoning → Send it to the Brain."
• "This is a routine task → Send it to a Specialist."

📚 The Librarian | AI Search

What it does

The city's master librarian who knows where every piece of trusted information is stored and retrieves the right knowledge when needed.

What it is (for understanding)

• This component is all about knowledge retrieval—finding the right information.
• It's not generating new content; it's surfacing existing knowledge.
• It understands context to return the most relevant results.

Real-world example : A support agent asks "What's our policy on refund requests?"

The Librarian searches the knowledge base, policies, and instance data to retrieve the exact information needed.

💬 The City Guide | Virtual Agent

What it does

The friendly chatbot that welcomes citizens through a conversational interface, understands their natural language needs, and connects them with the right AI services or information.

What it is (for understanding)

• This is your conversational AI chatbot (NAVA)—the interface users interact with directly.
• It engages in natural conversation, not just keyword matching.
• It understands context and intent from what users type in conversational language.

Real-world example : A user types in the chat: "I can't log into my account."

The Virtual Agent responds conversationally:

"I can help! Let me check a few things. Can you tell me what error message you're seeing?"

The user responds with additional details.

The Virtual Agent either resolves the issue directly in the chat (such as a password reset or account unlock) or routes the request to the appropriate support team with all the necessary context already gathered.

👨‍🔧 The Specialists | Now Assist Skills

What it does

The city's expert professionals, each trained to perform a specific task such as writing, summarizing, analyzing, or generating content.

What it is (for understanding)

• Each Specialist is purpose-built for a specific domain task.
• One Specialist might excel at writing, while another specializes in code analysis.
• Think of a city with electricians, plumbers, architects, and engineers—each with their own expertise.

Real-world example : Now Assist Skills include

• Writing Specialist – Drafts emails and documentation.
• Analyzer Specialist – Examines data and identifies patterns.
• Summarizer Specialist – Condenses long pieces of text.

🤖 The Problem Solvers | AI Agents

What it does

The city's intelligent workers who take ownership of a request, make decisions, use available tools, and complete tasks from start to finish.

What it is (for understanding)

• These agents are autonomous but governed.
• They can make multi-step decisions and use multiple tools.
• They don't just respond—they act and complete tasks end-to-end.

Real-world example:  A user asks: "Create an incident, assign it to the right team, and send them a notification."

An AI Agent breaks the request into steps, uses the appropriate APIs and tools, makes routing decisions, and completes the entire workflow without human handoff.

✈️ The Overseer | AI Control Tower

What it does

The city's operations control center that monitors all AI activities, coordinates workflows, and ensures everything runs smoothly and reliably.

What it is (for understanding)

• This is your governance and orchestration layer.
• It sees and coordinates all AI activity across the city.
• Think of an airport control tower ensuring all planes land safely and on time.

Real-world example: Multiple AI requests are happening simultaneously—some from the Librarian and others from the Problem Solvers.

The Overseer ensures they don't interfere with one another, tracks resource usage, logs activities for compliance, and flags any issues.

🛡️ The Protector | Guardian

What it does

The city's security force that safeguards every interaction by blocking harmful content, protecting sensitive information, and enforcing AI safety policies.

What it is (for understanding)

• This runs in parallel with inference—not as an afterthought.
• It detects and blocks harmful outputs, data leaks, and policy violations.
• Think of it as both a security officer and a content moderator.

Real-world example : A user submits a request containing sensitive financial information or attempts to misuse the system.

The Protector detects the sensitive content, masks or redacts it where appropriate, and ensures AI safety policies are enforced.

🧠 The Brain | Now LLM Service

What it does

The city's central intelligence that understands language, reasons through problems, and generates human-like responses for all AI services.

What it is (for understanding)

• This is the core generative AI capability—language understanding and reasoning.
• It powers much of the city's intelligence.

Key point: The Brain is not exposed directly to customers. It works behind the scenes through other AI components such as the Specialists and Problem Solvers.

Multiple options available:

• Now LLM 
• Azure OpenAI
• Google Gemini
• Anthropic Claude 

Real-world example: When a Specialist needs to write an email or a Problem Solver needs to reason through business logic, they call upon the Brain, which generates intelligent, contextual responses.

🌉 The Bridge Builder | AI Agent Fabric

What it does

The city's network of roads and bridges that connects ServiceNow AI with external systems, applications, and AI agents, enabling seamless collaboration across organizations.

What it is (for understanding)

• It allows ServiceNow AI to securely connect with external systems.
• Think of it as the highway system connecting your city to neighboring cities.
• It enables your AI agents to communicate with customer systems, partner platforms, and third-party APIs.

Real-world example : An AI Agent needs to retrieve customer information from an external database or call a third-party API.

The Bridge Builder enables that connection securely and efficiently, allowing the AI Agent to complete its work.