|
New for this update: |
|
Upgraded LLM |
|
Improved response structure, with increased accuracy in generated answers. |
|
Expanded Alix Knowledge Base with AKMS, j5, CESARII, Databridge Pro, Python Studio product documentation. |
|
Strengthened ethical and policy guardrails to define clear boundaries on content policies, prevent bias, protect privacy and enforce transparency requirements. |
|
Guided responses with verified sources - Questions can be directed to either the general knowledge base (i.e., the LLM’s training data) or the Hexagon product knowledge bases. Alix will display a visual indicator showing the source of each answer, enabling users to trust authoritative Hexagon information while also asking general questions with confidence, minimizing the risk of hallucinated responses. |