DataQRL
Enterprise data governance platform — 7 modules, real-time integrity, entity resolution at scale.
What was broken.
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.
How it's wired.
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.
What I shipped.
- ▸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
What changed.
Why these choices.
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.