Everything behind Remedi, from the cloud to the people who support it.
Remedi is hosted on Google Cloud, reasons with Anthropic’s Claude and records every step on Restate. It is built to HIPAA and SOC 2 standards, connects to the systems you already have, and is supported by the team that builds it.
- Google Cloud
- Anthropic
- Restate
- Cloudflare
- PostgreSQL
Each piece does one job, and does it well.
Google Cloud
Hosting
Remedi’s servers, database and secrets live on Google Cloud in the United States. Google Cloud holds SOC 2, ISO 27001 and HITRUST certifications and offers HIPAA-eligible services.
Anthropic
Reasoning
Anthropic’s Claude reads each request, weighs your rules and drafts the explanation your team sees. Your data never trains it.
Restate
Durable execution
Restate records each step as it happens. If a server restarts mid-task, the work resumes where it stopped, and nothing is done twice.
Cloudflare
The front door
Cloudflare delivers every page and screens every request, with a firewall and protection against floods of traffic.
PostgreSQL
The record
One relational record of every schedule, shift, rule and decision, encrypted and backed up with point-in-time recovery.
Host Remedi in our cloud, in yours, or on your own servers.
The same Remedi wherever it lives. Options marked on request are set up for your organization during onboarding.
Remedi’s cloud
We host, patch and monitor Remedi on Google Cloud in the United States, around the clock.
Your cloud
Remedi deployed in your own AWS, Microsoft Azure or Google Cloud account, so your data stays inside your security boundary.
Your data center
Remedi on your own servers or private cloud. It ships as containers, PostgreSQL and Restate, so nothing ties it to one vendor.
Claude does the reasoning, and you choose where the model lives.
Every model works under the same rules, wherever it is hosted. Each action is logged with its rule, its reason and the model version. Patient information goes only to a provider under a business associate agreement, and your data never trains a model.
Claude, from Anthropic
The reasoning behind Remedi’s forecasts, fills and explanations. Lighter work goes to lighter, faster Claude models.
Claude in your cloud
The same Claude models through Google Vertex AI, Amazon Bedrock or Microsoft Azure, under your own cloud agreement.
Open-weight models
Open models for targeted work, hosted in your cloud or on your own hardware, so the model sits inside your walls.
Built to HIPAA and SOC 2 standards.
HIPAA
Business associate
We sign a business associate agreement before any protected health information enters scope, and we keep the HIPAA Security Rule’s administrative, physical and technical safeguards.
SOC 2
Type 1 audit in progress
Remedi is built to the SOC 2 criteria for security, availability and confidentiality, and our controls are monitored continuously in Drata.
In every deployment
- Encryption in transit and at rest, with AES-256 on every database and disk
- Multi-factor sign-in on every production system
- Role-based access, one login per person and no shared accounts
- Every read and write of patient information logged with who, what and when, and kept six years
- Patient identifiers removed from logs, error reports and exports by the code itself
- Each organization’s data walled off from every other’s, tested on every build
- Point-in-time backups and a recovery plan we test
- Hosted in the United States
- Your data never pooled, never sold and never used to train a model
- Export your data whenever you like, and have it returned or destroyed when you leave, confirmed in writing
For your security team, on request
- Single sign-on through Okta, Microsoft Entra ID or Google, by SAML or OpenID Connect
- Accounts that open and close automatically with your directory (SCIM)
- The audit log streamed to your security monitoring
- Encryption keys you hold and can revoke
- A private network connection to your data center for HL7 and file feeds
Our security overview, subprocessor list and completed questionnaire are shared under NDA. Write to [email protected].
Every change is checked before it reaches you.
Written in Rust
Remedi’s server is written in Rust, a memory-safe language that rules out whole classes of security flaws.
Checked before it ships
Every change passes formatting, static analysis, thousands of automated tests and a dependency security audit before it reaches you.
No standing keys
Deployments sign in with short-lived credentials, and every secret sits in a managed vault with each access logged.
Governed everywhere
Every action is logged with its rule, its reason and the model version, and your team can reverse any of them.
Remedi connects to the systems you already have.
It reads from them and writes back through the standards healthcare already uses, read-only first, until your team trusts the results.
- FHIR R4
- SMART on FHIR
- HL7 v2
- X12 270/271
- X12 837
- SAML
- OpenID Connect
- SCIM
Timekeeping and workforce systems
Timecards kept in sync in both directions.
HR and payroll
Rosters, positions, pay rules and hours worked.
Your EHR
Census and admissions through FHIR R4 and HL7 v2, and Remedi opened from inside the chart with SMART on FHIR.
Payers
Coverage checks through X12 270/271 and claims through X12 837.
Fax and documents
Referral packets and orders read from fax and scans, with a person confirming what was read.
Your directory
Single sign-on through your identity provider, with accounts that follow your HR record.
Your data warehouse
Your Remedi data delivered to Snowflake, BigQuery or Databricks.
Public records
CMS staffing records, the NPI registry, state license boards, the federal exclusion list, weather, and flu and RSV activity.
A senior team builds Remedi and supports it.
Behind the founders is a bench of engineers, scientists, clinicians and reviewers who have built software that health systems and millions of people depend on. Lighthouse partners work with them directly.
Meet the foundersEngineering leadership
Engineering leadership from enterprise security software sets how Remedi is built, secured and released.
Product and systems engineering
Engineers build and harden the services, the connections and the apps your team uses.
Governance review
A reviewer checks the deployed system against its documentation: versioned, monitored, auditable and measured against a baseline.
Workforce science and statistics
Researchers and university statisticians decide how results are measured before the results come in.
Clinical guidance
Clinicians keep each staffing decision grounded in how care is delivered.
Implementation and support
Our team sets Remedi up with yours and supports it, so you hire no one to manage it.
To evolve how healthcare works — so better care costs less and reaches more people.
We’re looking for the right Lighthouse Partners.
This isn’t a trial. It’s a partnership for the long term: we build Remedi with you on your own floors, your staff use it free for a year, and you keep a partner discount for life.