An interface between Wallaroo.ai and Equitus.us KGNn could be established by using Wallaroo.ai's workload orchestration features to process data from KGNn.
___________________________________________________________________________
Wallaroo.ai uses YAML (.yml
) files for configuration and installation, and ZIP files containing instructions, requirements, and artifacts for its workload orchestrations. Equitus.us's KGNn platform automatically ingests, unifies, and contextualizes data into an "AI-ready knowledge graph."
A possible workflow for integration would be as follows:
Data Ingestion and Contextualization: Equitus.us's KGNn would ingest and process disparate data, structuring it into a knowledge graph.
Wallaroo.ai Workload Orchestration: A Wallaroo.ai workload orchestration, defined by a ZIP file, could contain a Python script that connects to the Equitus.us KGNn API. This script would fetch the "AI-ready" data for a specific inference task.
Inference Execution: The Wallaroo.ai inference engine would then execute the model on the data retrieved from KGNn.
Result Handling: The orchestration could be configured to send the inference results back to Equitus.us to be integrated into the knowledge graph, or to another specified destination.
Wallaroo.ai's API and SDK offer a lightweight way to make these connections, allowing for custom data sources like Equitus.us KGNn to be integrated into Wallaroo.ai's automated and scheduled ML workflows.