Introducing AnswerRocket's Latest Updates: Enhanced Interface, Expanded Deployments, and More!
We are excited to announce a new update for AnswerRocket, bringing a host of powerful new features and improvements to enhance your experience and streamline your workflows. Here's a detailed look at what's new:
1. Power Chat User Interface
- Larger Content View: Enjoy a larger view of any content generated by Max, keeping your chat conversations clean and focused.
2. Deploy Max in Azure
- Managed Application Deployment: You can now deploy Max into your own Azure Cloud instance as a Managed Application, offering greater ease, flexibility and control over your environment.
3. Skill Studio Improvements
- Enhanced Chat Playground: Access the chat playground directly from within Skill Studio and easily swap models to quickly experiment.
- SQL Playground: Write SQL directly from a data connection within Max, making it easier to explore your data connections.
- LangSmith Environment Override: Override the LangSmith environment used for tracing at the Copilot level.
- Support for Anthrop\cs Models: Max now supports Anthrop\cs models, broadening your options.
- Control Skill Output Downloads: Skill developers can now choose between downloading facts in raw format to Excel or as the formatted HTML tables shown to the user.
4. Diagnostic Improvements
- Focused Diagnostics View: When viewing diagnostics, only the executed steps in the pipeline are shown, simplifying the user interface.
- User Experience Enhancements: Small improvements and bug fixes enhance the overall user experience.
5. SDK Improvements to Chat ask_question()
- Model Override: Override the model used by the chat pipeline for question execution.
- Single Skill Constrainment: Constrain the chat pipeline to use only a single skill instead of all skills within a copilot.
- Custom Message History: Pass in custom message history with your chat question for more precise context management.
6. Chat Pipeline Enhancements
- Prompt Engineering for Suggestions: Max can now provide suggestions directly from the chat pipeline. By designing a prompt that includes
<suggestion>{suggestion_text}</suggestion>, the suggestion_text will be automatically converted to a chat suggestion.
These updates are designed to empower you with more control, flexibility, and efficiency in your interactions with Max and the AnswerRocket platform.
Tested Models
OpenAI
Chat: gpt-4o-2024-05-13
Narrative: gpt-4o-2024-05-13
Evaluations: gpt-4o-2024-05-13
Embedding: text-embedding-3-small
Gemini
Chat: gemini-1.5-pro-latest
Narrative: gemini-1.5-pro-latest
Evaluations: gemini-1.5-pro-latest
Embedding: models/embedding-001
Anthrop\c
Chat: claude-3-5-sonnet-20240620
Narrative: claude-3-5-sonnet-20240620
Evaluations: claude-3-5-sonnet-20240620
Embedding: NA
Updated