RKB109's picture
Publish artifacts for knowledge-graph-risk-engine-20260719
66485a3 verified
Raw
History Blame Contribute Delete
1.82 kB
{
"name": "Knowledge Graph Risk Engine",
"problem": "Risk teams need relationship-level explanations instead of opaque entity scores.",
"domain": "knowledge-graphs",
"architecture": "graph",
"hugging_face_tasks": [
"token-classification",
"feature-extraction",
"question-answering",
"sentence-similarity"
],
"recommended_stack": [
"FastAPI for entity and evidence APIs",
"Neo4j Community or PostgreSQL recursive queries",
"Sentence Transformers for entity resolution",
"NetworkX for local graph validation",
"Kafka-compatible event ingestion",
"OpenTelemetry for lineage and query traces"
],
"real_world_data_sources": [
{
"name": "SEC EDGAR submissions API",
"url": "https://data.sec.gov/submissions/CIK0000320193.json",
"purpose": "Public company and filing relationships"
},
{
"name": "GLEIF LEI API",
"url": "https://api.gleif.org/api/v1/lei-records?page[size]=5",
"purpose": "Public legal-entity identifiers and relationships"
}
],
"job_description_skills": [
"Entity resolution and relation extraction",
"Knowledge-graph modeling and path queries",
"Graph-based explainability and provenance",
"Streaming ingestion and schema evolution",
"Risk-model evaluation and data quality controls"
],
"impact_targets": [
"Reach entity-resolution precision >= 0.95 on reviewed pairs",
"Return evidence paths for 100% of emitted risk flags",
"Process 10,000 relationship events per minute in load tests",
"Detect schema and orphan-node regressions in CI"
],
"baseline_evaluation": {
"test_examples": 4,
"accuracy": 1,
"synthetic_evaluation": true
},
"estimated_delivery": "8-12 weeks for one engineer",
"generated_baseline_is_production_ready": false
}