Overview of Zuora AI Assistant
Zuora’s AI Assistant for Analytics is an AI-enabled conversational copilot. It helps you find real-time, business-ready metrics quickly and efficiently. The AI assistant’s conversational interface fosters productivity by enabling you to access your business data no matter where it resides. It eliminates the barriers of dependency with analysts or developers. You can interact with the AI Assistant in natural language to ask data-related questions.
The AI Assistant offers text-based responses with visualizations and tabular data for comprehensive insights, based on the complexity of your question.
Zuora’s AI Assistant is designed to meet industry standards. Yet, before you use the AI Assistant, be aware of the following technical limitations:
- The AI Assistant’s responses might include inaccurate data. We recommend you review and evaluate the responses for accuracy before using them.
- The AI Assistant feature currently works best in the English language only.
Business benefits
Zuora’s AI Assistant provides these main capabilities to support you in your journey to data discovery:
- Explore data easily
You can quickly access your business data in the natural language regardless of your expertise in complex SQL. Zuora’s AI Assistant understands and interprets natural language and responds in easy-to-read narratives accompanied by visualizations and data tables that you can click.
- Get actionable insights
Real-time visualizations, tabular data, and text-based answers help you identify trends, anomalies, and patterns that impact their subscription and billing operations.
- Expedite work
With conversational AI, you can get immediate real-time metrics in response to your questions, which enables faster decision-making and operational efficiency.
- Gain complete visibility
The AI-driven analytics experience provides a comprehensive view of subscription and billing metrics, which helps you gain a holistic understanding of their business performance.
- Democratize data access
Using a natural language interface, Zuora’s AI Assistant democratizes data access and analysis for users with varying levels of technical expertise. Effortless interaction with data enables a wider range of stakeholders to derive value from analytics and make data-informed decisions.
- Monitor and improve accuracy
To understand how you interact with Zuora’s AI Assistant, the system collects feedback and usage patterns to continuously improve the accuracy and relevance of the generated responses. Over time, the insights will be more consistent and accurate, which can enhance the overall user experience.
Key features of the AI Assistant for Analytics
Zuora’s AI Assistant has the following key features:
- Seamless integration
The AI Assistant seamlessly integrates with analytical data sources, transactional objects, reports, and data queries, which enhances the accuracy and speed of delivering answers.
- Drilldown for deeper insights
You can apply drill-down on all summable metrics and a few non-summable metrics to gain more insights within the dataset.
Technical constraints of Zuora AI Assistant
During the Beta phase of Zuora’s AI Assistant, only a limited set of analytics-related metrics can be supported. This limitation is temporary and specific to our early Beta phase.
- Time period: Because of internal limitations with the large language model (LLM), we can only reliably support up to 2 years of data per query. However, queries on data beyond two years might still work but might not be accurate.
- Asynchronous time periods are not supported (gaps) ( "show me the net MRR for June 2022 to July 2023, August 2022 to June 2023"): Time periods must be synchronous and continuous. Pick between two intervals (e.g show me the net MRR between June 2022 to July 2023)
- Future time is not supported: We process present events only. Asking for a future time will return an error
- Spelling and grammar: Natural language can vary greatly between individuals and might cause ambiguity. The queries entered might include the use of colloquial language, slang, or non-standardized grammar, which poses more challenges to the NLP algorithms to interpret your context and intent as a user. For the most accurate results, construct queries without any misspellings or grammatical errors.
- Filtering (e.g., where clause): Because of the current limitation in our LLM, we can only reliably support filtering for selected attributes. To know more, see Get Started with Zuora AI Assistant.