DataQRL

Enterprise data teams lose weeks reconciling records across systems. Existing MDM tools are heavyweight, opaque, and require expensive consultants. Mid-market companies need governance that scales with them — not against them.

Event-driven core. DataPulse streams change-data-capture into a validation engine (Sentinel). ReconFlow runs Splink-based probabilistic matching with hierarchical fidelity scoring. SignalWatch wraps anomaly detection on top. Cortex orchestrates AI agents over the lake; CaseHub gives investigators a cockpit; Data Studio handles remediation. All seven modules share a typed contract layer.

• ▸DataPulse™ — real-time integrity monitoring over CDC streams
• ▸ReconFlow™ — distributed entity resolution & data lineage (Splink + custom scoring)
• ▸SignalWatch™ — anomaly detection with adaptive thresholds
• ▸Cortex™ AI orchestrator + Sentinel validation engine
• ▸Data Studio remediation workstation & CaseHub investigative cockpit
• ▸Closed-loop verification processing millions of records

Chose Splink over Zingg for explainability. Picked Neo4j for lineage graphs (vs JSON in Postgres) once depth > 4. Kept the agent layer behind a strict tool interface so swap from OpenAI to local models is one config change.