URL Caching for Cost Savings
Gathering votes
When a user pastes in a public URL
- you download
- convert to PDF
- train the model
Potentially that could be cached for cost savings AND performance.
The cache could be updated when the page has changed
The cache could expire when not requested again within x hours/days
The only thing private is that user x requests page y
But the page itself is public and would have run through multiple caches on the way to you.
Could be a separate internal service that doesnt even know about the user and just returns a model based on a URL.
Just some food for thought to support your great product
Patrick
2 months ago
One vote
Categories
Feature