INTEGRATION
Charles Trinh — Oracle Health (Cerner) Integration | Las Vegas, NV · Las Vegas, Nevada · Henderson · Clark County, Nevada · North Las Vegas, Nevada

Oracle Health carries two platforms under one brand, and an integration that ignores the split will hit a workflow the current FHIR surface does not cover. I design for both surfaces and retire the legacy bridge on a date.
The two surfaces you actually have to integrate with
Oracle Health is, for the next few years, two platforms wearing one brand. The newer FHIR surface is what clients see in the developer portal and what current documentation describes. The Millennium stack underneath still exposes many operational workflows only through legacy interfaces: the HL7v2 TCP connections, the CernerCcl database objects and the older CCDA export paths. A build that targets only the FHIR surface will eventually hit a workflow that exists only on the legacy side, and a build that targets only the legacy side is investing in a platform Oracle is actively retiring.
My rule is to integrate against the FHIR surface for every read and write that the current documentation covers, and to bridge to Millennium only for the workflows that the customer's operations depend on and the FHIR surface does not yet expose. Each bridge I build is scoped to a small set of legacy interfaces and documented as a transitional path. When Oracle Health brings the corresponding workflow into the FHIR surface, the bridge is retired. The bridge is not a permanent architecture choice; it is a planned debt that I size and date.
The bridge debt has a budget. Every Millennium interface I bridge is approximately one engineer-week of build plus a maintenance load of one to two days a quarter for code-set drift. I write that cost into the plan and the customer's finance team sees it before the build starts. Bridges that nobody budgets for become the things that break quietly on a Saturday morning when the customer has forgotten they exist.
Authentication and the tenant model
Oracle Health's modern authentication uses OAuth 2.0 with SMART on FHIR scopes, similar to the Epic model, but the tenant identifier is different and the customer's security office owns a separate allow-list. getting it wrong produces a plausibly-shaped error that says the application is not registered. I confirm the tenant identifier against the application's developer portal entry before the first test call, because a wrong tenant ID can waste a week of the customer's security review.
Service accounts in Oracle Health are provisioned per environment and per tenant. The non-production environment is a separate tenant with its own identifiers, and the token minted for production does not authenticate against the non-production tenant at all. This is more forgiving than Epic, where the mirror is a refresh of production, but it introduces a different trap: the non-production tenant may be running an older FHIR surface with a different set of operations. I test the same call against production and non-production in one working session, and record the difference in the build's README, because the gap is never zero.
Pass-through authentication is the variant I see when the customer wants their own clinicians to authorise on the application's behalf. Oracle Health supports this through SMART launch contexts, but the launch context in the non-production tenant does not include the patient the clinician can see in production. Testing a pass-through workflow in non-production requires a specific patient that is provisioned for testing, and the customer's operations team knows which one. I ask for it during the kick-off, not during the test window.
The FHIR surface and what it does not do
The Oracle Health FHIR surface implements the US Core profiles on the resources it exposes, and conformance is declared in the CapabilityStatement. I read that statement before writing any code. It is the only reliable source of which operations are supported and where Oracle Health has chosen a narrower interpretation than the published profile. Anything the statement marks as not supported is a candidate for a Millennium bridge or a different design.
The most common gap I see is search parameters. Oracle Health supports a subset of the search parameters on the resources it exposes, and a query that works on one FHIR server will frequently return a 400 on this one because a parameter is missing or has a different cardinality. I build a parameter matrix for each resource the application uses, run each parameter once against the tenant, and record the result. A developer who adds a search parameter without checking gets a failing test before review.
The second gap is write operations. Oracle Health accepts writes on fewer resources than US Core requires for the ONC (g)(10) patient access requirement, because the requirement is cast around reads from the patient's perspective. If your application wants to write Observation or DocumentReference, you need to check the statement, and often you need to confirm the write workflow with the customer's operations team. Writing is not a protocol decision here; it is an operational one, and the customer's owner decides which write paths the application is allowed to take.
Code sets, terminology trees and the quiet drift
Oracle Health maintains its own terminology trees on top of SNOMED CT, LOINC and ICD-10-CM. The customer's organisation can add local codes, and those local codes appear in the FHIR surface as values that map to a local code system, not to the standard. The application has to map these correctly, and the mapping table has to live where the customer can maintain it. Any mapping that lives only in the application's code becomes a monthly incident.
The drift I see is the local code that quietly gains and loses synonyms. A term that used to resolve to a LOINC code is now a local code, and an application that hardcoded the LOINC value stops matching. The symptom is a report with missing rows and no error message. I build the mapping table with version history and a weekly check against the CodeSystem resource, so any change to a local code set shows up as a diff before it reaches production. The check is cheap; the incident it prevents is not.
The mapping table is also the artefact the customer's terminology team wants to own. I hand it over with a short guide and a review cadence, and I ask them to review it whenever their Epic upgrade or their terms-of-use for the code set changes. In Oracle Health I have seen a code system's licence terms change quietly with a platform update, and the customer's compliance team needed to know within a month. The handover includes the trigger list for that review.
Millennium bridges and when to build them
A Millennium bridge is warranted when the workflow is operational, the customer cannot replace it, and there is no path on the FHIR surface within a quarter. The classic examples are scheduling, results distribution to a small clinic network, and a lab interface that the customer's reference lab only speaks in HL7v2. I refuse to bridge anything that is only used by an analytics report, because the analytics team can wait for the FHIR surface to mature and the bridge is not worth the maintenance.
The bridge uses the HL7v2 socket with the customer's legacy credentials and a translation layer that produces current FHIR resources for the downstream consumers. The translation layer is the part that has to be tested carefully. HL7v2 messages carry optional segments, and Oracle Health's Millennium sends some fields differently from what the interface specification document describes because of local customisation. I test the translation layer against a captured sample, not the sample in the specification, and I store the sample in the customer's environment so the tests can be re-run when the customer upgrades.
Every bridge has a deprecation date. The date is the one Oracle Health's product roadmap gives for the corresponding FHIR operation, rounded up to the end of the quarter. At the date, I run a full comparison between the bridge and the new FHIR path, switch the downstream consumers, and switch off the bridge. The customer has a short notice window. This is the discipline that keeps the bridge debt from becoming the architecture.
The acceptance test and who signs it
A live Oracle Health interface has two signatures: the customer's operations owner and the customer's security owner. The operations owner signs that the workflow produces the right result in front of a clinician. The security owner signs that the calls are made with the agreed scopes and the audit trail shows each call. Neither signature is redundant. The operations owner has caught builds that were technically wired but visually wrong in the clinician's screen, and the security owner has caught builds that leaked patient context across contexts during a refresh.
My acceptance test is a written script the two owners run against the non-production tenant. Each step records the FHIR call and its response, and each step is checked against the CapabilityStatement and the scope map. The script and its responses go into the ticket that closes the engagement. In a typical Oracle Health build I hand over forty to sixty steps; the smallest I have signed had eleven, and the largest ran past two hundred because the customer wanted a full regression list.
The FHIR facade work is where these bridges get a cleaner home when the downstream consumers are multiple. The facade can present a FHIR surface that combines the current Oracle Health surface and the legacy interfaces, and the customer gets a single endpoint with a single contract. If you want the broader picture, the HL7 v2 interface work and the interoperability programme describe how these pieces fit into a larger estate.
Proof
- ONC (g)(10) self-test PASS, 317 of 317 subtests, April 2026 — public suite URL
- 67 public hostnames, 51 serving, one AWS account
- 104 clearinghouse enrolments across 53 jurisdictions
- USCDI v7 proposal — Patient Digital Identity Credential — listed at isp.healthit.gov
- SAM.gov UEI MVKQW9DKKGB8 · CAGE 20KM5 · NPI 1194645564
- AWS Business Associate Addendum active since 18 August 2026
Where this work is delivered
Engagements run from Las Vegas and reach across the valley: clients in Las Vegas, NV; Las Vegas, Nevada; Henderson; Clark County, Nevada; and North Las Vegas, Nevada — plus remote delivery nationwide.
How an engagement runs
1. Trace one real case
Follow a single patient or claim end to end with a timestamp at every hop, before any document is written.
2. Written findings
A findings list naming the defect, the system, the owner and the date.
3. Monthly review
Review the list monthly and check whether the dates held.
Frequently asked
What is this service?
This page describes charles trinh — oracle health (cerner) integration — the scope, the failure modes it prevents and how the work is measured. It sits in the Interoperability category at trinh.io.
Where is the work delivered?
Engagements run from Las Vegas, Nevada across the valley — Las Vegas, Henderson, North Las Vegas, Spring Valley, Paradise, Winchester, Sunrise Manor, Enterprise, Summerlin, Clark County — and remotely anywhere in the United States. On-site work in the Las Vegas valley, remote delivery elsewhere.
What does it cost?
Standard rate $195 per hour, committed $165 per hour at twenty hours or more per week for three months or more; $1,600 standard day, $1,350 committed day. Scoped in writing before work starts, fixed fee where the deliverable is defined.
Who is Charles Trinh?
A technology architect in Las Vegas, Nevada, working in healthcare identity, consent and interoperability. Not a physician, not a medical practice, and not affiliated with any other Charles Trinh.
How does an engagement start?
With one trace of a real case and a written findings list naming the defect, the system, the owner and the date. The first call is twenty minutes and free: 702-559-3489.
Engagement
Scoped in writing before work starts; fixed fee where the deliverable is defined. First call is 20 minutes and free. 702-559-3489 · charles@trinh.io
Talk to Charles
- Phone: 702-559-3489
- Email: charles@trinh.io
- Based in Las Vegas, Nevada · working remotely across the United States
Record
Profile
Healthcare technology architect working where federal regulation becomes production infrastructure. Built and operates a FHIR R4 clinical platform — SMART App Launch authorization server, patient and provider portals, SQLite — on a hardened single-host AWS estate of 67 public hostnames. The platform returned a self-test PASS on the ONC (g)(10) Standardized API test suite in April 2026, 317 of 317 subtests, publicly verifiable. Designed a consent architecture in which a permissioned ledger records patient grant, refusal and revocation as immutable events while all clinical data stays off-chain and encrypted per patient — no PHI on the ledger by design. Filed a proposed USCDI v7 data element with ONC on verifiable patient identity. Translates the Federal Register, HHS/CMS guidance and HITRUST requirements into architecture, controls and audit evidence.
Selected engineering scale
- Self-managed estate of 67 public hostnames, 51 serving, on one AWS account.
- FHIR R4 API with a SMART App Launch authorization server; patient and provider portals; an AI service layer; Hyperledger Fabric peer, orderer and CA; IPFS; a 337-table SQLite clinical store.
- Two permissioned distributed ledgers in production, one operating a consent registry that records grant, refusal and revocation with no PHI on chain.
- X12 005010 pipelines across eight transaction sets.
- Designed, built, deployed and operated end to end by one person.
Credentials and standards work
- Inferno (g)(10) self-test PASS, 317 of 317 subtests, April 2026 — the ONC Standardized API test suite. Standards validated: US Core 6.1.0, USCDI v3, SMART App Launch 2.0.0, Bulk Data 2.0.0. Public suite URL.
- USCDI v7 public comment submission, April 2026 — proposed Patient Digital Identity Credential under the Provenance data class, listed publicly on the ONC Interoperability Standards Platform.
- SAM.gov UEI MVKQW9DKKGB8 · CAGE 20KM5 · NPI 1194645564. AWS Business Associate Addendum active 18 August 2026.
Technical environment
AWS us-east-1 (compute, storage, messaging, SMS, email, voice) · Linux and systemd hardening · nginx · Fabric and EVM ledgers · FHIR R4 · X12 005010 · HL7 · SQLite · self-hosted mail, drive, search and monitoring · Git-based change control and CI release gates.