AEGENT-lite

AEGENT-lite is an AI agent built for Microsoft 365 Copilot that gives security and engineering teams fast, dependable answers to everyday cyber & AI security questions, using a Kainos-maintained knowledge base of standards, frameworks and best practice.

Ask AEGENT-lite what a control means, what a framework requires, or how a common practice compares to a standard, and get a straight answer with a source from our authoritative knowledge base, whilst also surfacing and summarizing your organization's connected Microsoft 365 knowledge sources without leaving Microsoft Teams.

Key benefits

• Faster answers: You get a sourced answer to a common security question in seconds instead of digging through documents or waiting on a consultant.
• Consistent answers: You get the same grounded answer every time, based on the same maintained knowledge base, not whoever happens to be free to ask.
• No cost to start: You can install it free from Microsoft Marketplace and try it with your team before deciding on anything paid (the same solution forms part of our paid-for AEGENT.security platform).

Intended audiences

• Security and GRC teams who need quick answers on frameworks, controls and best practice.
• Security architects and CISOs evaluating Kainos's security products.

Features

• Configurable access to your Microsoft 365 knowledge sources (SharePoint, OneDrive, Email, Teams, People and selected web pages).
• Grounded in a Kainos-maintained knowledge base covering standards and frameworks such as NIST CSF and SOC 2.
• Conversational Q&A on security controls, compliance frameworks and general best practices.
• Sourced citations on every answer, so you can see exactly where the guidance came from.
• Runs inside Microsoft Teams through Copilot, with no separate app to install or maintain.

Use cases

• Company policy guidance: Provides clear summaries of current security policies, standards, guidelines, and codes of conduct.
• Project initiation assistant: Offers essential guidance, requirements, and expectations for new projects based on contracts and scoping documents.
• Framework guidance: Explains what a specific control in a framework like NIST CSF or SOC 2 actually requires.
• Policy sense-check: Gives a quick read on whether a general practice lines up with common industry standards.
• Audit prep: Refreshes what evidence a given control typically needs before a review.
• Terminology and onboarding: Helps less experienced team members understand security terms and controls.

Support

For setup or support, contact us: support@aegent.security

Legal disclaimer

Full terms of use are available here

This product is an AI-powered agent that uses large language models and retrieval techniques to generate responses and recommendations based on a knowledge base of best practices. The Agent may generate inaccurate, incomplete, or outdated outputs. Outputs are advisory only and should be reviewed by a human before taking actions.

Kainos has designed the agent in accordance with responsible AI principles including transparency, accountability, privacy and security.

Kainos has assessed this agent as not being a high-risk AI System under Annex III of the EU AI Act. To the extent any customer use may cause the deployment to fall within any high risk category, the customer shall be solely responsible for assessing its own deployment.

Prerequisites, limitations, and conditions

• Requires a Microsoft 365 Copilot licence.
• Answers are grounded in general security knowledge, plus the material you as a user have access to within your organisation.
• Answers are general guidance, not a substitute for a formal audit, legal advice, or a Kainos security consultant's assessment.
• You must maintain human review of all outputs and remain solely responsible for decisions taking using the agent.
• Accuracy depends on the currency of the underlying knowledge base, which Kainos maintains and updates, and the information available within your own organisation.
• The agent must not be used in the following ways:
  • as the sole basis for cyber & ai security or other decisions;
  • for automated legal or other determinations;
  • for medical diagnosis;
  • for activities prohibited by AI regulation or other applicable laws.

Data processing

As the agent has been built for Microsoft 365 Copilot, the following items shall be governed by the customer's arrangements with Microsoft for Copilot:

• The extent to which the agent can access information made available through Microsoft Graph and customer-configured repositories;
• Retention of customer prompts, no training commitments, and data residency arrangements.