User Guide: Data Workbench
1. Overview
The Data Workbench is the central repository for managing the information ChartGen AI uses to answer your questions. It serves as a unified hub where you can access industry-standard public datasets or securely manage your own private data sources.
2. Public Data
"Ready-to-use datasets for immediate insight."
The Public Data tab offers a library of curated, high-quality datasets.
- Instant Access: These datasets are pre-processed and ready for analysis immediately—no cleaning required.
- Fusion Analysis: This is ideal for benchmarking. You can combine these external standard datasets with your internal private data to perform comparative analysis (e.g., comparing your company's Private Sales Data against Public Economic Indicators).
3. Private Data
"Your secure data hub."
The Private Data tab is where you manage your own organizational data. There are two ways to bring data into ChartGen AI:
3.1. File Upload (Local Data)
For ad-hoc analysis or static reports:
- Click "+ Add Data" or "Connect Data".
- Drag and Drop your files into the upload area.
- Supported Formats:
.csv,.xls,.xlsx. - Constraint Note: You can upload up to 10 files at a time (max 100MB per file).
3.2. Data Integrations (Live Connections)
For continuous monitoring, connect directly to your business applications. ChartGen AI supports:
- Finance & Ops: QuickBooks, NetSuite, Stripe, eBay.
- Databases: MySQL, PostgreSQL, SQL Server, Snowflake, Azure SQL.
- And more. Feel free to contact us if you need to connect other data sources.
Once connected, ChartGen AI can pull fresh data automatically, ensuring your reports always reflect the latest numbers.
4. Smart Semantic
"Transform raw files into analytical models."
Simply uploading a spreadsheet isn't enough for deep AI reasoning. The Smart Semantic engine is a powerful feature that structures your data for accurate analysis.
How it Works
When you run the "Smart Semantic" tool:
- Transformation: ChartGen AI scans the raw data to identify logical relationships.
- Definition: It automatically extracts and defines:
- Metrics: Quantifiable numbers (e.g., Revenue, Churn Rate).
- Dimensions: Categorical data (e.g., Region, Product Category, Time).
- Tables: How different data sheets relate to each other.
Why Use It?
- High Stability: It reduces AI "hallucinations" by grounding the model in structured definitions.
- Deep Analysis: It enables complex operations like Attribution Analysis (understanding what drives a change in a metric) which is impossible with flat, unstructured text.
5. Managing Your Workbench
In the Private Data list view, you can track the status of your assets:
- Search: Quickly find datasets by name.
- Status Tags: Look for the "Semantic" tag next to a filename. This indicates the data has been successfully processed by the Smart Semantic engine and is ready for deep analysis.
- Task History: Click this to view the log of semantic processing jobs.