Skip to main content

v1.64.0

Features & improvements:

  • New: Ability to compare settings between jobs and adjust multiple jobs at once
  • New: Ability to configure API Key permissions
  • New: Ability to select different preprocessors for decision points
  • New: Added API endpoints for managing Agentic Workflows
  • New: Arcanna Assistant now supports Retrieval-Augmented Generation (RAG) collections
  • New: Option to copy job index mapping when cloning a job
  • Improvement: Added a more friendly name for a session in Agentic Workflows
  • Improvement: Added a visual warning when one input integration is invalid on a job with multiple inputs
  • Improvement: Added metadata to Agentic Workflows
  • Improvement: Added search bar on Context Enrichment configuration when searching through available integrations
  • Improvement: Adjustments to the structure of internal metrics endpoints
  • Improvement: Arcanna Assistant now supports GPT-oss with Ollama
  • Improvement: Job Overview now offers more metrics

Bug fixes:

  • Fixed a corner case where "undecided consensus" filter would not work
  • Fixed an corner case where programatically updating jobs would break the auto-training feature
  • Fixed an issue where a naming conflict could appear when saving a new Context Enrichment
  • Fixed an issue where Arcanna's internal fields we're not available to be selected as decision points
  • Fixed an issue where Filters did not work when using external storage
  • Fixed an issue where filters using "empty_string" values would not work on RCA type of jobs
  • Fixed an issue where job counter would return wrong stats
  • Fixed an issue where renaming a conversation in Agentic Workflow would cause an error
  • Fixed an issue where some filter options would not work when using an external output integration instead of Arcanna's internal integration
  • Fixed an issue where the LLM component would not connect to Elasticsearch component
  • Fixed an issue with the "http timeout in seconds" parameter of Splunk integration
  • Similarity table required some adjustments
  • Tags required some adjustments for a better user experience