Find files by type, size and stored date.Find files by type, size and stored date.
Keldra indexes file metadata so your application can find the right objects by size, content type, committed time or path.
Narrow a large collection before downloading its files. Spend less time transferring payloads just to discover what they are.
A file collection with indexed metadata
Stored attributes let a query narrow the collection without reading every payload.
File
content_type
content_length
Committed
Path
report.pdf
application/pdf
1,482,113
2026-10-12
/docs/report.pdf
image.jpg
image/jpeg
3,921,408
2026-10-11
/images/image.jpg
notes.txt
text/plain
4,293
2026-10-10
/notes/notes.txt
data.csv
text/csv
892,204
2026-10-09
/data/data.csv
archive.zip
application/zip
14,201,883
2026-10-08
/backups/archive.zip
Filter stored attributes before downloading files
Before: application loops
Open files one by one to discover their format
Read or fetch attributes separately for every object
Repeat work across folders and collections
Assemble the final matching set in application code
Finding the right file becomes a slow, repetitive task.
After: indexed fields
Filter stored content types to find the right format
Use indexed file sizes to narrow the result set
Filter committed time to find recent or older files
Use path and version information to locate the object
Fetch the original files that actually match.
The metadata projection also includes content hashes. It does not automatically extract arbitrary EXIF, inspect every file format or repair a supplied content type. Use typed JSON or full-text indexing when those are the facts you need to search.