Documentation Index

Fetch the complete documentation index at: https://kb.kentik.com/llms.txt

Use this file to discover all available pages before exploring further.

February 2026 - AI / Insights

  • Published on Jun 29, 2026
  • 1 minute(s) read
  • Ben F
  • LZ
  • AC
 Prev Next

The following updates were released to the Kentik platform for AI / Insights during February 2026.

Features

  • AI Advisor Data Explorer integration: AI Advisor can now use your current Data Explorer query as context so you can reason about and modify queries directly, and "Send to Data Explorer" loads results instantly.
  • AI Advisor Synthetics subtest results: View subtest results filtered by specific agents and targets directly through AI Advisor, with automatic target resolution for URL, page load, hostname, and transaction tests.
  • AI Advisor device config tool: Improved device configuration tooling with fuzzy device-name search, company-wide config revision listing, in-config search, and date/time filtering.
  • AI Advisor message manager: Better context retention across longer conversations, fewer token-limit errors, and more stable performance in extended sessions.
  • AI Advisor alert tool: More reliable alert filtering with fuzzy policy matching as a fallback and NMS policy support.
  • AI Advisor AWS mapping: AI Advisor now maps and displays AWS VPC and Subnet names alongside their IDs.
  • AI Advisor CSV download: Added a "Download CSV" button to tables in the AI Advisor Journey feature.
  • AI Advisor tool results: Tool-call results with no results can now be opened and viewed in the detail pane, and tool-call details now expand by default.
  • AI investigation notifications & summaries: Investigations now send completion notifications to configured channels (email, Slack) and generate concise AI summaries for faster review.
  • Claude Opus 4.6 and Sonnet 4.6: New Claude Opus 4.6 and Sonnet 4.6 models are available through the AI Gateway, with Sonnet 4.6 now the default advanced tier.

Bug Fixes

  • AI Advisor response handling: Fixes an issue where Claude Sonnet/Opus occasionally returned raw Markdown instead of the expected JSON, causing errors.
  • AI Advisor performance: Optimized system prompt handling with caching and clearer section organization.
  • AI Advisor Overlay mode fixes: Deleting a session from the Overlay mode no longer causes unwanted navigation, and the left edge resize handle is more reliable.
  • AI Advisor duplicate alarms: Alarms with different dimensions are no longer incorrectly treated as duplicates.
  • JSON schema responses: The AI Gateway no longer drops response_format for requests that expect a json_schema response.
  • On-prem stability: Fixed a tiktoken cache issue that was breaking the AI Gateway in some on-premises deployments.
Related articles