Max.ai Analytics Assistant Release Notes - Version 25.02
Release Date: January 29, 2025

Introduction

The Max.ai Analytics Assistant version 25.02 introduces a robust suite of features aimed at improving diagnostics, refining user and developer workflows, and enhancing overall usability. This release prioritizes user feedback to deliver targeted upgrades to the chat, code, and survey functionalities.

Key Improvements

Enhanced Code Version Management

We have simplified the experience of managing versions of individual code blocks within your components. This allows for saving as you move along, and easily revert back to an earlier version if needed.

User Surveys

You can now survey your users on a regular interval to learn how the product is working for them and gather feedback. Learn more on how to set this up here.

Resource Links in Output Nodes

Resources such as images can now be referenced via direct links within output nodes. This allows skill outputs like dynamic layouts to directly access assistant and skill resources, making integration more seamless. So if your visual output requires an image you can now host that image within the Resources section of Max.

Additional Enhancements

Improved Diagnostic Panel

The diagnostic panel now displays the last input message by default along with the corresponding output message.
Users can expand an accordion to view the rest of the input messages, making it easier to trace issues without extra clicks.

Playground Enhancements

Improved text editing experience:

One can edit text areas without having to go into edit mode and hitting save. The text area is always available for editing.
Pressing Enter now inserts a line feed rather than triggering a save.

Context Appending for Artifacts Downloads

User chat questions are appended as context at the end of artifact download file names, ensuring better traceability.

Log Files Available Upon Node Completion

Addressed a major frustration for developers as now log files for individual nodes become accessible as soon as the node completes, without waiting for the entire skill to finish execution.

Surface Failures Due to Issues When Communicating with the LLM

Errors encountered in chat completion attempts appear directly in Max’s response message. This replaces the generic "contact support" default response with actionable error messages for better troubleshooting.

Tested Models

This release has been tested with the following models to ensure compatibility and stability:

OpenAI:

Chat: gpt-4o-2024-08-06
Narrative: gpt-4o-2024-08-06
Evaluations: gpt-4o-2024-08-06
Embedding: text-embedding-3-small

Gemini:

Chat: gemini-1.5-pro-002
Narrative: gemini-1.5-pro-002
Evaluations: gemini-1.5-pro-002
Embedding: models/embedding-001

Anthropic:

Chat: claude-3-5-sonnet-20241022
Narrative: claude-3-5-sonnet-20241022
Evaluations: claude-3-5-sonnet-20241022
Embedding: Not Applicable

Changelog