Match every record. Trust every result.
Purpose-built for HIPAA, GLBA, GDPR, and FedRAMP environments where a wrong match isn't just an error. It's a liability. Deploy in the cloud or fully on your own infrastructure.
Regulated industries can't afford false matches
In financial services, healthcare, and government, a single misidentified record can trigger fraud, regulatory penalties, or patient safety failures. Existing tools force a trade-off between speed and accuracy. Derive doesn't.
Compliance exposure
False positives in AML and KYC workflows trigger regulatory action under GLBA, GDPR, and FinCEN beneficial ownership rules. A wrong match is a liability, not just an inconvenience.
Rules-based fragility
Hand-crafted matching rules break with every new data source. Edge cases multiply faster than engineering can keep up, and every exception needs a new rule.
No explainability
Black-box systems can't tell regulators or auditors why two records were matched or rejected. In regulated industries, an unexplainable decision is an indefensible one.
Cloud lock-in
Most modern IR platforms are cloud-only. HIPAA, FedRAMP, and GDPR data residency requirements make public cloud deployments a hard compliance blocker for most regulated buyers.
Principle-based entity resolution: precise, explainable, deployable anywhere
Derive replaces brittle rules with three universal behavioral principles (frequency, exclusivity, and stability) that govern how every identity attribute is scored. No training. Easily tuned. No cloud dependency.
Principle-based matching
Deterministic and probabilistic scoring governed by attribute behavior. Works out of the box on any data domain with no pre-training required.
Full audit trail
Every match includes a confidence score, feature breakdown, and reasoning trace. Every decision is defensible to regulators, auditors, and legal teams, by design rather than by retrofit.
Flexible deployment
SDK-based architecture runs on-prem, private cloud, AWS, Azure, or fully air-gapped. Your data never leaves your infrastructure. No Derive personnel ever access your systems.
A $7B global market growing at 11% annually
The pure-play IR/ER market is driven by compliance mandates, data proliferation, and the failure of rules-based systems. Regulated industries (BFSI, healthcare, government) are the fastest-growing and highest-value segments.
Precision and deployment flexibility. Not one or the other.
Every alternative forces a trade-off. Cloud-native platforms can't meet on-prem mandates. MDM incumbents are expensive, slow, and require expert configuration. Derive delivers enterprise-grade accuracy with zero infrastructure compromise.
Derive
Quantexa
IBM Match 360
AWS Entity Resolution
Resolve every patient record and own the identity that results.
Derive links the same person across every source feeding your exchange, claims database, or health system, then keeps the resolved identifiers in your environment, tunable to your population. Purpose-built for HIEs, state APCDs, and the health systems that feed them.
One engine, tuned to how you resolve identity
The matching problem looks different from an exchange, a claims database, and a health system. Pick yours.
A full resolution pipeline, running on Spark
Derive is not a bolt-on matcher. It cleanses, blocks, matches, and clusters at scale, then hands you a persistent, governed identity base you keep and tune.
Preprocessing & cleansing
Standardizes and repairs messy source data (names, addresses, dates) before a single comparison runs.
Block analysis
Intelligent blocking keeps comparison tractable across billions of records without missing true matches.
Tunable matching rules
Configure matching to your own population and data quality, not a vendor's fixed, one-size model.
Transitive closure
Resolves full identity clusters across all matches, so A=B and B=C correctly yields one person, not three.
Customer-owned identifiers
Persistent unique IDs live in a dynamic knowledge base you own, never rented back from a vendor cloud.
Deploy where your data lives
Run it in your cloud or on-premise. PHI never has to leave the boundary you already control.
Every feed
EHRs, payers, labs, registries
Standardize
Repair & normalize
Tunable rules
At Spark scale
Cluster
One identity per person
Knowledge base
Persistent IDs, yours
The difference, point by point
Measured against the tools HIEs, APCDs, and health systems actually evaluate. Each wedge is tied to a documented difference, not a slogan.
| Capability | Derive | Referential (Verato) | Cloud meter (AWS ER) | Legacy EMPI (Rhapsody) |
|---|---|---|---|---|
| You own the resolved identifiers | Yes, in your environment | Vendor-hosted | Match results returned | Varies by deployment |
| Accuracy past the referential ceiling | Tunable + transitive closure | Capped ~95.8%1 | Rule / ML, fixed | Referential add-on |
| No dependence on a reference database | None required | Core dependency | N/A | Optional module |
| Cost when you reprocess often | Flat, unlimited runs | Subscription | Per-record, per run | Per-identifier |
| Runs fully on-premise | Yes | Cloud-only | Cloud-only | Yes |
| Scale of preprocessing | Spark, billions of records | Service-bound | Service-bound | Single-node tendency |
◆ You own the identity base
Persistent identifiers and the dynamic knowledge base live in your environment, re-usable forever and governed by you. Not a subscription you rent, and not match results handed back from someone else's cloud.
◆ Accuracy you can tune
Tunable rules plus transitive closure push match quality past the ceiling a referential-first approach runs into, because you're resolving your population rather than matching against a fixed national reference set.
◆ Predictable economics
A flat annual license with unlimited reprocessing. Your bill doesn't move when you re-link on every ingestion cycle, and it's the same whether you run in the cloud or on-premise.
◆ Data stays in your boundary
Deploy in your own cloud or on-premise so PHI and member data never leave the environment you already control and audit, a requirement for most exchanges and every state program.
Matching just became a measured, funded priority
The MATCH IT Act
A bipartisan Senate companion bill (Aug 2026) would create a Medicare bonus for accurate patient-match rates, turning match quality into a reimbursable, measured metric rather than a back-office cost.
TEFCA is live
With 11 QHINs, 10,600+ connected organizations, and 115M+ documents exchanged, nationwide exchange makes high-quality identity resolution infrastructure-critical.
Data sovereignty
Regulated and state programs increasingly require patient and member data to stay inside their own boundary, exactly where Derive runs and where cloud-only matchers can't follow.
Priced to be predictable
An annual license by record volume, with unlimited reprocessing and the same price in any environment. Start with a pilot; grow as your volume does.
License tiers
Record volume under management · unlimited reprocessingA record is an input source record under license for the term: counted once, re-matched as often as you like. No per-run metering, no re-buying your own identifiers.
See it resolve your data
Bring a de-identified sample from your own sources. In a working session we'll tune the rules to your population and show you the match rate you can actually reach.