Lux Overview

Follow

What Is Lux?

Lux is Kion’s in-app AI agent designed to help organizations understand, manage, and take action across their cloud environment more efficiently. Lux enables users to ask questions in natural language, receive actionable insights, and safely automate tasks, without requiring deep technical expertise or familiarity with Kion’s feature set.

Lux works directly within the Kion platform, using your connected cloud data, permissions, and approved APIs to provide context-aware guidance, reporting, and automation.

Managing cloud environments requires coordination across finance, security, operations, and engineering teams. Lux reduces friction by:

  • Turning questions into answers quickly (no dashboards to hunt through)
  • Guiding users toward next steps, not just raw data
  • Helping teams automate routine tasks safely, such as budget creation or updates
  • Reducing reliance on specialized cloud expertise for everyday decisions

Lux helps organizations move faster while maintaining control, transparency, and governance.

What Can Lux Do?

Highlights of Lux include:

  • Surfacing spend report data visually and providing analysis or recommendations on the data. It can also forecast based on your spend
  • Creating and implementing compliance checks; Lux contains a sub-agent tool allowing it to write actionable cloud custodian policies
  • Managing budgets, enforcements, labels, funding sources, and updating the org chart
  • Gathering data on compliance, users, and cloud access roles
  • Understanding and explaining Kion’s feature set via documentation
  • Chat and Conversation history is maintained

See our AI Prompt Library for examples of prompts for use with Lux, as well as our MCP.

Lux is currently offered at no additional charge with an active Kion license.

Key Benefits to Your Cloud Estate

  • Improved visibility into spend, compliance, and operational data
  • Faster decision-making through conversational access to insights
  • Safer automation with built-in approvals and guardrails
  • Broader adoption by supporting both technical and non-technical users
  • Flexible AI choice, allowing organizations to bring their preferred provider and model

Prerequisites

To use Lux, the following must be in place:

  • An active Kion environment
  • Cloud accounts onboarded and configured within Kion
  • An AI provider connected to Kion (see Providers in the next section)
  • Network access that allows communication with the selected AI provider

    • Lux LLM Provider Network Access Requirements The below outlines the network endpoints and ports that must be accessible for Lux to connect to LLM providers.

    Amazon Bedrock

    Endpoints:

    • Regional endpoint pattern: bedrock-runtime.{region}.amazonaws.com
    • Common regional endpoints:
      • bedrock-runtime.us-east-1.amazonaws.com
      • bedrock-runtime.us-west-2.amazonaws.com
      • bedrock-runtime.eu-central-1.amazonaws.com
      • bedrock-runtime.ap-southeast-1.amazonaws.com
      • bedrock-runtime.ap-northeast-1.amazonaws.com 
      • Port: 443 
      • Protocol: HTTPS

    Anthropic API

    Endpoint:api.anthropic.com 

    Port: 443 

    Protocol: HTTPS

    OpenAI API

    Endpoint:api.openai.com P

    ort: 443 

    Protocol: HTTPS

    Firewall Configuration

    Configure outbound firewall rules to allow TCP traffic on port 443 to the following destinations:

    | Provider | Destination | Port |

    | Amazon Bedrock | bedrock-runtime.*.amazonaws.com | 443 | 

    | Anthropic | api.anthropic.com | 443 | 

    | OpenAI | api.openai.com | 443 |

No additional infrastructure deployment is required.

Supported AI Providers and Models

Lux supports the following AI providers:

  • Anthropic
  • OpenAI
  • Amazon Bedrock

Organizations connect their own provider and model of choice.

Kion strongly encourages using the recommended providers and their models for Lux to ensure optimal performance and reliability. While other models may function, Kion cannot guarantee results outside of the recommended options.

Kion plans to add support for Google Gemini later in 2026.