Key Concepts¶
This page explains some key concepts around the Numbers Station platform to help you understand how to think about the product.
Datasets¶
A dataset is a collection of tables that all answer questions from one domain. This allows you to have a contextual conversation about a topic. In the datasets panel, the number of tables is denoted by the number on the bottom left of each dataset panel.
After clicking on a specific dataset to inspect it, you can access the knowledge related to that dataset. By clicking All tables and views you can choose a specific table to manually look at.
Conversations¶
A conversation is a collection of question and answer pairs, where the question is asked by the user, and the answer is generated by the AI agent. An answer can be text or a data result, which is the generated table after an SQL query is run. Context is the portion of a conversation before a question is asked. Conversations do not share context, which means that an AI agent cannot access a different conversation to answer a question with no context. Every new conversation is a clean slate.
Dimensions¶
Dimensions are categorical attributes that describe the data in your datasets. They provide the context for analysis, such as time periods, geographic locations, or product categories. Dimensions help in organizing and segmenting your data, making it easier to drill down into specific areas of interest and gain deeper insights. In the platform, dimensions are used to slice the data, enabling more detailed and granular analysis.
In the dataset inspection view, clicking on Dimensions allows you to access all dimensions and their relevant information.
Metrics¶
Metrics are quantitative measurements that are used to track and assess the performance or progress of specific business activities. They are often derived from numerical data and provide key insights into trends and patterns within the dataset. Examples of metrics include sales figures, conversion rates, and average order values. In Numbers Station, metrics are calculated and displayed to help users understand the underlying data and make informed decisions.
In the dataset inspection view, clicking on Metrics allows you to access all metrics and their relevant information.
Filters¶
Filters allow users to refine and narrow down their data analysis by including or excluding specific data points based on defined criteria. By applying filters, users can focus on particular subsets of data, such as a specific time frame, geographical area, or product category. This helps in isolating relevant information and performing more targeted analyses. Filters enhance the precision and relevance of the data insights generated by the platform.
In the dataset inspection view, clicking on Filters allows you to access all filters and their relevant information.