Unstructured data: the missing link in AI context engineering
Impossible d'ajouter des articles
Échec de l’élimination de la liste d'envies.
Impossible de suivre le podcast
Impossible de ne plus suivre le podcast
Unstructured data: the missing link in AI context engineering
-
Lu par :
-
De :
Komprise has amassed $85 million in funding and multiple patents for its metadatabase, which lends structure to unstructured data. Data processing is increasingly in the spotlight as enterprise AI struggles to reach production.
"It's garbage in, garbage out," according to Krishna Subramanian, COO at Komprise, something more enterprises are realizing as they develop context engineering practices for AI agents. But Subramanian argues that common techniques such as RAG and vectorizing data aren't enough to ensure quality results.
In this episode, we also discuss:
- Metadatabases vs distributed data lakehouses
- Metadatabases vs knowledge graphs
- The art and science of data classification
- Why, at Komprise, "we don't believe in 100% automation."
And more!
References:
- Context is king as agents evolve, but problems persist
- Komprise adds Iceberg for unstructured data management
- Atlassian's Secoda acquisition adds data catalog for Rovo AI
- ServiceNow digs deeper into data for AI with Data.world buy
To learn more about software development and platform engineering, check out Search IT Operations.
To watch clips from our podcast, subscribe to our YouTube channel, @EyeOnTech.
