Akeyless AI Insights

Overview

Akeyless AI Insights enables natural-language interaction with the Akeyless platform using Large Language Models (LLMs). To operate, AI Insights must be configured at:

Account level - feature enablement
Gateway level - specify LLM target + model

Supported LLM Providers

OpenAI (GPT models)

Prerequisites

Before you begin, ensure you have:

Akeyless CLI installed & authenticated (admin access)\
LLM Provider account + API Key
- OpenAI — https://platform.openai.com/api-keys
Akeyless Gateway running
Ability to create and manage Targets in Akeyless
Protection key available (for encrypting API credentials)

High-Level Steps

Step	Description	Tool
1	Enable AI Insights at account level	CLI
2	Create OpenAI Target	CLI
3	Configure Gateway for AI Insights	REST API
4	Validate configuration & test	CLI / UI

Step 1 - Enable AI Insights (Account Level)

akeyless update-account-settings \
  --enable-ai-insights true

To verify:

akeyless get-account-settings

Expected result contains:

"ai_insights": { "enable": true }

To disable:

akeyless update-account-settings --enable-ai-insights false

Step 2 - Create LLM Target

Option A - OpenAI Target

Command

akeyless target-create-openai \
  --name <target-name> \
  --api-key <openai-api-key> \
  [--openai-url <base-url>] \
  [--model <default-model>] \
  [--organization-id <org-id>] \
  [--key <protection-key>]

Example

akeyless target-create-openai \
  --name my-openai-target \
  --api-key sk-xxxx \
  --model gpt-4

Find Target ID

akeyless get-target --name <target-name>

Model Rules

Provider	Valid Prefix
OpenAI	`gpt-`

Examples:

gpt-4, gpt-3.5-turbo
gpt4

Step 3 - Configure Gateway

Configuration uses the Gateway API.

TOKEN=$(akeyless auth | grep token | awk '{print $2}')

curl -X PUT "http://localhost:8000/config/ai-insights" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $TOKEN" \
  -d '{
    "cluster_identity": {
      "account_id": "a-1234567890",
      "access_id": "p-1234567890",
      "cluster_name": "my-gateway"
    },
    "ai_insights": {
      "enable": true,
      "target_name": "my-openai-target",
      "model": "gpt-4"
    }
  }'

To disable:

"ai_insights": { "enable": false }

Verification

Verify Account Setting

akeyless get-account-settings

Verify Target

akeyless get-target --name my-openai-target

Verify Gateway Config

curl -X GET http://localhost:8000/config/ai-insights

Test in UI

Open Akeyless Web UI
Navigate to AI Insights
Start a chat
Ask a natural language question

Troubleshooting

Issue	Resolution
`AI Insights disabled`	Enable at account level
`Gateway disabled`	Update gateway config
Invalid model	Must use `gpt-`
Invalid target	Must be OpenAI
Target not found	Validate target name/ID
Authentication failure	Re-auth with `akeyless auth`
Gateway unreachable	Check port 8000 + firewall
API key errors	Check validity + base URLs

Configuration Checklist

Step	Status
Enable AI Insights	☐
Create Target	☐
Store Target ID	☐
Configure Gateway	☐
Verify Gateway Config	☐
Test in UI	☐

Updated about 4 hours ago