GraphSentry

GraphSentry is a web app (with optional CLI agent) that continuously monitors and “tests” your knowledge graph for drift, broken assumptions, and low-quality data. It connects to Neo4j, Amazon Neptune, or RDF triple stores and runs scheduled checks: schema/ontology conformance, orphaned nodes, duplicate entities, stale facts, and suspicious relationship spikes. It also tracks provenance and freshness, flags sources that are degrading, and generates a weekly “graph health” report that product, data, and KM teams can actually act on. An AI layer helps propose fixes (merge candidates, property normalization, missing links) and drafts Jira tickets with evidence, but it does not auto-write to production by default. The product is intentionally narrow: it’s not another graph database, not another visualization tool, and not a generic data quality platform—it’s a specialized QA and observability layer for knowledge graphs.