Skip to main content

GPTfy 1.7 Release Notes

Release Date: Jun 30, 2024

Enhancements​

  • Where Clause Validation : 'Where Clause' on the prompt is now validated at activation time.
  • Ability to Order By Any Field : Any field can now order related records on the Data Context Mapping. This is useful for extracting specific records. This enhancement solves requirements like extracting the latest 5 Opportunities based on the Expected Close Date. Users can Order By the Expected Close Date and set a Limit of 5.
  • Malformed JSON Response : Malformed JSON can cause unexpected errors (e.g., Code 60). It is difficult to troubleshoot these errors as malformed JSON can cause a fatal error. This release has made changes to capture the malformed JSON in the Exception Log and Security Audit records. Additionally, this exception has been handled so that end users will see clear error messages on the GPTfy User Console.
  • Conditionally Generate Blank JSON Nodes : Customers sometimes need the Data context-based extraction to show blank JSON nodes. This is especially true for customers considering Einstein Bot integration with GPTfy. The default Organization Level Value, 'Blank Replacement Value,' has been added to the 'GPTfy Settings.' Any Value populated here is used as the value for Blank Nodes. Customers must use this functionality only if recommended by the GPTfy success team.

Issues Fixed​

  • API Data Source : Callout of API Data Source inside of AI Mass processing was causing Mass processing jobs to fail. This has been resolved.
  • GPTfy User Console : Some of the action icons like Send Email, Create Task, and ThumbsDown were not available to End user through the GPTfy User Console. This was caused by missing permissions in the 'GPTfy User' permission set. This permission set has been updated.

Known Issues​

  • Gmail Plugin File Processing : Prompts with 'Include Files' checked do not work on GPTfy Gmail plugin.
  • JSON Prompt Timestamp Field : Timestamp field on Prompt record does not work for JSON prompts.
  • Pre-installed Data Context Mapping : Data Context Mapping created at the time of GPTfy installation has the status of Inactive.
  • Azure RAG Activation : Azure RAG Activation process fails. Customers exploring Azure RAG should reach out to GPTfy Success team.
  • AI Model Card Count : AI Model card on GPTfy Cockpit sometimes shows incorrect count.
  • AI Model Import : AI Model Import sometimes creates duplicate records.