ADVISORY

Charles Trinh — Platform Reliability Engineering | Las Vegas, NV · Las Vegas, Nevada · Henderson · Clark County, Nevada · North Las Vegas, Nevada

Reliability dashboard showing service objectives, error budgets and incident trends
Charles Trinh — Platform Reliability Engineering

Reliability in a clinical or payment platform is defined by patient and payer outcomes, not by uptime percentages. I build the service objectives, on-call model, error budgets and runbooks that let a team hold a reliability target and prove it.

Reliability has to be defined in clinical and financial terms

An uptime target of 99.9 per cent tells a patient nothing and a board very little. Reliability for a healthcare platform is defined in terms the operation understands: the percentage of eligibility checks answered in under two seconds during business hours, the percentage of claim submissions accepted on first send, the percentage of patient messages that reach the patient's chosen channel on the same day. I work with the operations lead and the clinical users to define those outcomes and then the service objectives that support them.

Each service objective has a stated measurement, a target and a window. The measurement is taken from the user's perspective, not from the server. A page that loads in the browser but cannot see an empty result is not a success; a check that returns a claim decision with a four-sided error is not a success. In my experience the failure is rarely where the team believes it is; it is in the entitlement check, the attribution table, or a match that returned nothing and logged nothing, and the measurement must be built so those show up.

The objectives are reviewed quarterly with the operation and the engineering team. If an objective is too easy, it is raised; if it is impossible given the dependencies, it is revised or the dependency is fixed. The review is not a negotiation about the percentage; it is a conversation about which user outcome is not being met and why. That framing keeps the reliability work coupled to the operation rather than drifting into technology for its own sake.

Error budgets make the trade

An error budget is the amount of unreliability the target permits before the team has to stop adding features and start improving reliability. If the eligibility objective is 99.5 per cent over a thirty-day window, the team has 0.5 per cent of the window to spend on incidents, planned maintenance and failed changes. When the budget is exhausted, the next sprint goes to reliability work. That policy is agreed with the business owner in advance, with a written route for exceptions, because without it the first exhausted budget produces a fight.

The policy is written in the operations language and approved by the head of operations, not only by the CTO. When the budget is spent, the business owner knows that new work will pause unless they override, and the override itself is recorded. In a healthcare platform the override conversation is usually uncomfortable and it is exactly the conversation the business owner needs to have. Without the policy, the override happens silently and the team pays for it in the next incident.

I link the error budget to the release process. When the budget is healthy, the release calendar runs normally. When it is low, the release gate adds additional approval steps and the change windows narrow. The link makes the reliability policy real at the moment it matters, which is when someone is about to ship a feature on a platform that is already shaky.

On-call that people can survive

On-call is the reliability practice that most directly destroys engineering teams when it is done badly. I design the rotation around the page rate the team can bear, the escalation paths and the support available out of hours. A rotation that wakes an engineer every night is not a rotation, it is an unpaid night shift with a pager. The service objectives and the alerting thresholds should mean that a page is rare enough that the rotation is sustainable, and where it is not, the engineering work is to reduce the page rate, not to add more people to the rotation.

Every alert is tied to a user-facing symptom and a response. Alerts that fire on a threshold nobody has looked at in six months are noise and get removed. I run an alert review monthly with the on-call engineer and the operations lead. The review looks at every alert that fired, what it was for and whether it resulted in action. The findings are recorded in the runbooks, and the alert list shrinks until it covers what the operations lead would actually wake someone for.

The escalation path is written, with names and numbers and a defined lead who takes over when the primary is unavailable. The observability and synthetic monitoring around the platform gives the on-call engineer the evidence they need to diagnose quickly. Good monitoring is not about the dashboards; it is about the five minutes between alert and diagnosis, and the runbook is where that time is spent or saved.

Runbooks as living documents

Runbooks are the operational memory of a platform and are the first thing to go stale. I write a runbook for every critical service and every recurring incident, with the steps, the commands and the verification that the fix has worked. Each runbook is exercised by an engineer who has not written it, before the service goes live, and again at least twice a year. A runbook that has not been exercised is a hypothesis, not a plan. The cost of discovering that at 3 a.m. on a Saturday is very high.

I keep the runbooks in the same repository as the code, with the deployment, the rollback and the incident response alongside each other. A change to the deployment updates the runbook in the same pull request, which is the only mechanism I have found that keeps the runbook close enough to trust. The inventory of runbooks is reviewed with the service catalogue quarterly and any runbook whose service has changed without a corresponding update is flagged.

When a platform migrates to a new environment, such as a healthcare cloud migration, the runbooks go through a full review before the cutover and a post-cutover update after the first real incident. That practice is how the new platform inherits the operational discipline of the old one instead of starting from a blank page that the team learns to fear.

Incidents, reviews and the honest blameless record

Every material incident gets a written review with a timeline, the contributing factors, the missed signals and the corrective actions with owners and dates. The review is blameless. It names roles and actions, not individuals, and it asks what the system allowed rather than who did it. That does not mean a person who made a mistake is shielded from a conversation; it means the record of the incident is about the system, so the team can talk about it and the organisation can learn. A blameful review produces a team that hides the next one.

The corrective actions are entered in the same tracking tool as the rest of the work, with an owner and a due date. They are reviewed monthly alongside the error budget. The engineering team's contribution to reliability is visible in the incident history, the action closure and the trend in the objective measurements. The operations lead gets a one-page monthly report on those trends, and the report includes what the team did not manage to fix, with the reason.

The reliability work I do for MyRxWallet includes a 51-hostname AWS estate and runs on exactly these practices: service objectives, error budgets, on-call and runbooks that get exercised. The same practices apply when a platform is small, provided the objectives and the on-call are proportional to the scale. The aim is a platform where the team knows how reliable it is, what the cost of failure is and where the next improvement sits, and can prove it from the evidence.

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 — platform reliability engineering — the scope, the failure modes it prevents and how the work is measured. It sits in the Security and Compliance 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

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.

TRINH — Charles Trinh, technology architect
TRINH

Charles Trinh

Fractional CTO/CIO and Technology Architect — Healthcare Data, Identity & Regulated Infrastructure · Las Vegas, NV

Las Vegas, Nevada · US citizen · Fractional and project engagements

(702) 559-3489 Email Charles Trinh