I’m working on uploading large datasets covering multiple brands into Amazon Quick (specifically into Spaces), but I’m running into file size limitations when uploading through the Quick site directly. The data volume exceeds what the UI allows in a single upload.
Does anyone know of a workaround or alternative method to get larger files into Spaces?
Hi @knoplizz, there are a few workarounds for the file size limitation when uploading to Spaces:
1. Split the file into smaller chunks
Break your large dataset by brand (or by another logical segment) so each file falls under the upload limit. Spaces can hold multiple files and your agent will query across all of them seamlessly. The experience is the same as having one large file.
2. Use a SPICE Dataset instead of file uploads
If your data is structured/tabular, consider using the Dataset resource type within your Space. SPICE datasets are designed for larger data volumes and support SQL queries. You can also set up scheduled refreshes on the SPICE side, which helps with ongoing maintenance.
3. Reduce file size before uploading
- Remove unused columns or redundant data
- Convert from Excel to CSV (smaller footprint)
- Compress date/number formats
- Filter to only the rows your agent actually needs
4. Multiple smaller uploads into the same Space
You can upload several files into a single Space. They don’t need to be consolidated into one. The agent queries across all resources in the Space, so splitting by brand or time period works well without any loss of functionality.
The most common approach is option 1, splitting by brand. It’s the least effort and doesn’t require rethinking your data architecture. If you have dozens of brands, option 2 (SPICE dataset) might be a better long term fit since it handles larger volumes natively.