-
Type: Improvement
-
Status: Resolved
-
Priority: Detail
-
Resolution: Fixed
-
Affects Version/s: None
-
Fix Version/s: NXP-9.10-HF23, NXP-10.3
-
Component/s: ML Data Enrichment
-
Release Notes Summary:Provides a generic caching system to enrichment process so that paid services are not called more than once without need.
-
Epic Link:
-
Sprint:nxAI Sprint 10.3.9, nxAI Sprint 10.10.1
-
Story Points:3
Enrichment services usually call paid external services, eg. AWS Comprehend. I noticed from the translation implementation that a cache would be useful to avoid unnecessary calls. e.g. the following call could be cached.
["{'source':'en','target':'es','text':'I am very disappointed','translated':'Estoy muy decepcionado.'}"]
If we implement a generic cache in AbstractEnrichmentService, any enrichment implementation could benefit from it.
- is related to
-
AICORE-56 Add AWS translate as an Enrichment Service
- Resolved