Requirement: Building a system to provide insights and information to external users in a natural language conversation format. The context revolves around querying numerical information as we have a table T containing the cashflow data for some clients.
The system can be is accessible via this website.
For developer technical development click here
For simplity, the database only have one table for storing the following fields:
| Field | Description |
|---|---|
| Store_ID | The unique identifier for each store |
| Store_Address | The physical address or location of each store |
| State | The state of each store. |
| Company | The name of the company that owns the stores |
| Company_ID | The unique identifier for the company associated with each store |
| Revenue | The income generated by each store |
| Expenses | The costs incurred by the store |
| Date | The date where the data is recorded |
| Client_ID | The unique identifier of the client |
There are two clients in this mock database. Client 1 and Client 2. Client 1 has access to the cash flow data of Coles and Woolworths, whereas Client 2 has access to the cash flow data of Kmart and Aldi.
- QA Module: Answer general queries abou the system
- Data Module: Query the cash flow data based on the client's input
- Forecast Module: Forecast future cash flow data based on the client's input
- Switch Module: Determine the appropriate module for the the client's input.
- How much income did I get last month?
- How much did I spend in Apr-2023?
- How will my cashflow look like in the next three months?
- What can you help me with?
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent