AI Assistant Overview

The AI Assistant is your on-demand infrastructure expert, accessible from any page in the Guardian Pro console. Ask questions about your AWS environment in plain English, and get immediate, context-aware answers backed by real-time data from your infrastructure.

Unlike a generic chatbot, the AI Assistant has direct access to your resource inventory, security findings, cost data, compliance status, and architecture topology. It does not guess -- it queries your actual environment and provides answers grounded in real data.

What the AI Assistant Can Do

The AI Assistant covers the full breadth of Guardian Pro's capabilities:

Capability	Examples
Infrastructure Health	"What is my overall health score?" / "Why did my health score drop?"
Security Analysis	"What are my most critical security issues?" / "Explain the S3 public access finding"
Cost Insights	"How much am I spending on EC2?" / "What are my top cost reduction opportunities?"
Compliance	"Am I compliant with CIS benchmarks?" / "Which SOC 2 controls are failing?"
Architecture	"What are the biggest risks in my architecture?" / "What happens if my primary database fails?"
Resource Search	"How many Lambda functions do I have?" / "Show me all unencrypted RDS instances"
Remediation	"How do I fix the RDS encryption finding?" / "Can Guardian Pro fix this automatically?"

The AI Assistant retrieves live data to answer your questions, so responses always reflect the current state of your environment.

Accessing the AI Assistant

The AI Assistant is available as a chat panel that can be opened from anywhere in the Guardian Pro console:

1. Click the chat icon in the bottom-right corner of any page.
2. The chat panel slides open, ready for your question.
3. Type your question and press Enter or click Send.

The panel remains open as you navigate between pages, so you can reference the assistant's answers while working in other areas of the console.

tip

Use the keyboard shortcut to quickly toggle the chat panel open and closed without reaching for the mouse. Press Escape to close the panel at any time.

How It Works

When you ask a question, the AI Assistant follows a multi-step process:

1. Intent classification -- Your question is analysed to determine what type of information you need (security, cost, architecture, compliance, or general).
2. Context gathering -- Relevant data is retrieved from your environment based on the question type. This may include active findings, cost trends, compliance scores, or resource configurations.
3. Analysis -- The AI processes your question alongside the retrieved context to formulate an accurate, specific answer.
4. Response -- The answer streams to your chat panel in real time, often with actionable recommendations and navigation links.

The entire process typically takes 3-10 seconds, depending on the complexity of the query and the amount of data retrieved.

Tool-Based Analysis

For complex questions, the AI Assistant uses specialised analysis tools behind the scenes. When a tool is active, you see a status indicator in the chat panel -- for example, "Analysing security findings..." or "Checking compliance status..." -- so you know the assistant is working on your answer.

These tools give the assistant access to the same detailed data that powers Guardian Pro's dashboards and reports, ensuring consistent and accurate answers.

Real-Time Streaming

Responses stream to the chat panel in real time, so you start reading the answer immediately rather than waiting for the entire response to be generated. This is especially useful for longer, more detailed answers.

Action Buttons

Many responses include interactive action buttons that let you take immediate action without leaving the chat:

• Navigate -- Jump directly to a relevant page, such as the Action Centre filtered to a specific finding.
• Remediate -- Trigger a remediation workflow for a specific finding.
• Run Scan -- Start a new infrastructure scan.
• View Details -- Open a detailed view of a resource, finding, or risk.

See Actions and Navigation for a complete guide to action buttons.

Conversation Context

The AI Assistant maintains conversation history, so you can ask follow-up questions that build on previous answers:

You: "What are my critical security findings?"

Assistant: "You have 3 critical findings: ..."

You: "Tell me more about the first one"

Assistant: "The S3 public access finding affects bucket..."

You: "Can you fix it?"

Assistant: "Yes, Guardian Pro can remediate this automatically. [Fix Now]"

The assistant remembers what you discussed earlier in the conversation, so you do not need to repeat context. See Conversation History and Management for more details.

Multi-Account Awareness

The AI Assistant is aware of which AWS account you are currently viewing. When you switch accounts using the account selector in the navigation bar, the assistant automatically adjusts its context to the selected account.

If you are viewing the aggregate organization view, the assistant can answer questions across all your accounts.

info

When you switch accounts, the AI Assistant starts a fresh context to avoid mixing data between accounts. Your conversation history for each account is preserved separately.

Subscription Tiers and Usage

The AI Assistant is available on all Guardian Pro subscription tiers. Each tier includes a daily usage allowance:

Tier	Daily Allowance
Starter	Standard daily limit
Business	Higher daily limit
Enterprise	Highest daily limit

Your remaining daily usage is tracked automatically. If you reach your limit, the assistant lets you know and resets at the start of the next day.

note

The daily usage limit applies to AI-powered responses. Browsing conversation history and using action buttons do not count toward your limit.

Privacy and Security

• The AI Assistant only accesses data within your Guardian Pro account and connected AWS accounts.
• Conversations are stored securely and are only visible to users within your organisation.
• The assistant cannot make changes to your AWS environment without your explicit confirmation (via action buttons).
• No customer data is used to train or improve AI models.

Next Steps

• Starting Conversations -- Learn how to get the most out of your first interaction.
• What to Ask -- Browse example questions organised by category.
• Actions and Navigation -- Understand the interactive action buttons in responses.
• Conversation History -- Manage your conversations and revisit previous answers.