livegraphs-django/TODO.md

LiveGraphs Project TODO

Dashboard UI Improvements

Responsiveness

• Fix dashboard graphs scaling/adjustment when zooming (currently requires page refresh)

Theming

• Add dark mode/light mode toggle
• Fix dark mode implementation issues:
  • Make charts display properly in dark mode
  • Fix the footer not changing color in dark mode
  • Adjust the sidebar nav-link styling for dark mode
  • Make the navbar have a different background color from the body in dark mode
  • Make theme toggle automatically detect and respect the user's system preference
  • Fix inconsistency between system dark mode preference and manual toggle
  • Ensure charts properly update in both scenarios (system preference and manual toggle)
• Implement smooth theme transitions
• Add Notso AI branding elements
• Implement responsive table design (reduce rows to fit screen)

Data Export

• Implement multi-format export functionality
  • CSV format
  • Excel format
  • JSON format
  • XML format
  • HTML format
  • PDF format
• Create dropdown menu for export options
• Make export data section collapsible (folded by default)
• Add company name, date and timestamp to exported filenames
• Update data view to show maximum 10 rows by default, with a "Show more" button to expand to 50 rows, or "Show all" to display all rows

Admin Interface Enhancements

Company Management

• Add company logo upload functionality
• Add direct CSV download button for each company (superusers only)
  • Include company name, date and timestamp in filename
• Add UI for customizing CSV column names

Data Integration

External Data Sources

• Implement periodic data download from external API
  • Source: https://proto.notso.ai/jumbo/chats
  • Authentication: Basic Auth
  • Credentials: [stored securely]
  • An example of the data structure can be found in jumbo.csv
    • The file that the endpoint returns is a CSV file, but the file is not a standard CSV file. It has a different structure and format:
    • The header row is missing, it is supposed to be session_id,start_time,end_time,ip_address,country,language,messages_sent,sentiment,escalated,forwarded_hr,full_transcript,avg_response_time,tokens,tokens_eur,category,initial_msg,user_rating
  • The coupling of endpoint to the company and the authentication method should be handled in the backend and the superuser should be able to change it.
  • The data should be stored in the database and the dashboard should be updated with the new data.
  • The csv also contains a column with full_transcript, which is a uri to a txt file, encoded in utf-8. The txt file is a raw transcript of the chat.
    • The txt file should be downloaded, parsed and stored in the database.
    • An example of such txt file can be found in 132f3a8c-3ba5-4d89-ae04-cd83f1bc5272.txt
    • Note that the User and Assistant messages can be multiline and can contain html, which should be safely handled, and if safe, rendered in the frontend.
• Add scheduling options for data refresh
• Add UI button to trigger manual data refresh

Technical Debt

Performance Optimization

• Profile and optimize dashboard rendering
• Implement lazy loading for dashboard elements

Testing

• Add unit tests for export functionality
• Add integration tests for data import process