AI is going to change healthcare, but only if clinicians can trust it and developers can build it without reinventing the plumbing every time. FHIR Agent Studio is my attempt at both: a place to build AI agents for FHIR workflows from reusable building blocks, run them on a real InterSystems IRIS for Health backend, and inspect exactly what each one did at every step.

It is my entry in the InterSystems AI Agents for FHIR contest, and it ships with twelve working agents over a synthetic FHIR repository of around 1,000 patients and 20,000 resources.

1. The End of "Pay-and-Chase": A New Era of Payment Integrity

For decades, healthcare payers have been trapped in a reactive cycle known as "pay-and-chase." Traditional payment integrity relies on retrospective audits—reviewing claims months after disbursement to claw back overpayments. This inefficiency costs the industry billions in unrecovered fraud, waste, and abuse.

ClaimAuditAI introduces a fundamental paradigm shift by moving the audit to the point of submission.

The successful construction and implementation of AI agents to address diverse use cases in the healthcare sector depend on high-quality data and APIs, effective governance, and management. The InterSystems IRIS FHIR server delivers all of this and is also fluent in Python, Vectors, and Interoperability. Combined with a strong LLM, patients, physicians, caregivers, and managers gain access to state-of-the-art technology for personal and public health.

Hi folks!

Suppose a new version of your amazing software solution built on IRIS needs some persistence change: some data needs to be converted, some classes get new properties or even renamed, and some data should be prepopulated for newly introduced persistent classes to keep the whole application data consistent.

What are your approaches? 

SQL-based scripts? ObjectScript, or Python? for both directions of migration, up and down? 

Do you autogenerate it with Claude/Codex? What about data-consistency tests?

COuld you please share your thoughts? or libraries?

Hi!

I'm trying to compiling a class with an Embedded Python and got this error:

Compilation started on 06/12/2026 11:44:17 with qualifiers 'cuk'
Compiling class User.VectorSearch
ERROR #7802: Worker job/s '749:33' unexpectedly shut down in group '#Default:(446070926892):0'.
ERROR #7812: Work queue unexpectedly removed, shutting down.
ERROR #5002: ObjectScript error: <THROW>WaitForComplete+215^%SYS.WorkQueueMgr *%Exception.StatusException ERROR #7802: Worker job/s '749:33' unexpectedly shut down in group '#Default:(446070926892):0'.
ERROR #7812: Work queue unexpectedly removed, shutting down.
Detected 3 errors during compilation in 1.020s.

 Hello Community,

I am learning more about InterSystems IRIS administration and would like to understand the recommended approach for monitoring performance in a production environment.

What tools, metrics, or dashboards do you typically use to monitor system health, database activity, and application performance?

I am particularly interested in learning which indicators are most useful for identifying potential bottlenecks before they affect users.

Thank you for sharing your experience and recommendations.

The CCR EHR Application Specialist Exam is now available on the InterSystems exam store! 

This exam certifies that developers are able to use Tier 2 CCRs to document and progress change workflows.

Most "AI agent + FHIR" projects end up with the same shape: a FHIR server over here, a vector database over there, and a Python service in the middle whose job is to call an embedding API, marshal float arrays back and forth, and keep two datastores in sync. Three moving parts, two network hops, and an embedding client you now own forever.

Triage Park: our entry for the InterSystems Programming Contest: AI Agents for FHIR, doesn't have any of that. The agent never computes an embedding. It never imports an OpenAI embeddings client. There is no vector database.

Hi All,

There is a steady increase in the amount of metadata that I need to build in the message flow (across BPs & DTLs) and I am interested on a good pattern to follow that allows for a mix of metadata types (e.g. json, xml list, etc) and is accessible quickly either within DTLs or BP.

One example for this would be adding various localised codesets to metadata during DTLs which would later be used to submit a FHIR translate request to a terminology server. Providing visibility in the DTL to which codesets are to be translated while batching up the translate request for performance.

Hi Developers,

We are happy to announce the new InterSystems online programming contest:

🏆 InterSystems Programming Contest: AI Agents for FHIR🏆

Duration: May 25 - June 14, 2026

Prize pool: $12,000

Virginia's got BIGDATA in IRIS and Iowa wants it

This distraction, "Meshing IrisClusters" with Cilium ClusterMesh, provides database access using the Enterprise Cache Protocol (ECP) from the Eastern Seaboard of the United States to the MidWest on Google Cloud Platform.  Ridiculously powerful setup for national/multi-national setups and though the use cases are a plenty, this particular one is a simple example on how to send ETL/ELT the unemployment office.

The rise of distributed systems and cloud-native architectures within large institutions has redefined hosting for InterSystems IRIS interoperability applications (productions). In modern environments, the standard layout involves at least two production instances operating on distinct nodes under a load-balancing configuration to handle requests.

The benefits are clear: you can now scale dynamically based on request volume, increasing the number of active productions during peak demand and going down when traffic subsides.

Overview 

This article explains how InterSystems IRIS Data Platform handles both transactional and analytic workloads without sacrificing performance. It also covers how the platform manages memory and scales under peak load, what its security model looks like, and how global banks and brokerages are using it in production today.

InterSystems IRIS Data Platform for Financial Services

InterSystems IRIS data platform is a technology for mission-critical financial services applications.

I would like to create a generic error handler defined somewhere in a class (or if not possible a routine that calls a class). 
This should not be page specific but for the whole app (as I have lots of CSP pages inside a legacy app). Obvisouly this will not be a single try catch block but rather something setup at system level (or web application level if not possible). The same way it's possible to handle sessions in a global way (by having aclass that inherit from %CSP.SessionEvents).

Hi everyone, 

I created a short demo and reference repository showing how to configure Auth0 as an OAuth provider for an InterSystems IRIS FHIR server, and how to call protected FHIR resources from a Python Flask application. 

The links below include a video walkthrough and a github repo with the sample application used in the video 

Links 

Video walkthrough: 

GitHub repository: https://github.com/MarshyKid/iris-fhir-oauth-demo

1. Introduction

Epidemiological surveillance is one of the foundational pillars of public health. Régis Júnior et al. (2026) define it as a continuous system of data collection, analysis, interpretation and dissemination of health events — a function whose effectiveness depends critically on the quality of information systems, data analysis capacity, and coordination between different levels of care.

I know IRIS has a built-in testing framework, but I'm not sure how to write tests that validate the full message flow through a Production without depending on external systems. Do you mock Business Operations? Is there a pattern you use to test integration scenarios in an isolated and repeatable way?

I'm trying to lock down access in our IRIS system.
I notice there is no current %DB_X resource (which I want to add to add a read-only role for it).
If I add resource %DB_X will there be an immediate effect or will it only be applicable to roles (and users linked to that role)?

I understand that Business Rules can be used to route or transform messages without touching code. But I'm not clear on when it's better to use a Business Rule versus handling the logic directly inside a BPL or a custom BP class. What criteria do you use to decide between them?

If I need to add or change properties in a message class that is already being used by a running Production, what is the safest way to do this? Are there risks with existing messages in the queue or in the message store? Is there a migration strategy the community recommends?

FHIR REST APIs are powerful and standardized, but they are not a general-purpose query engine.

They are great for operations such as reading resources, searching by supported parameters, following links, and working with Bundles. But many real-world questions require orchestration across multiple resources.

For example:

• Show diabetic patients with A1c above 9 in the last 6 months.
• Find diabetic patients whose HbA1c is getting worse despite active insulin therapy.
• Find patients with abnormal HbA1c and no follow-up encounter within 90 days.
• Find patients with overlapping anticoagulant and NSAID prescriptions.
• Build an evidence timeline for a patient.
• Explain why this patient matched the query.

Hi folks,

One of the things I've been thinking about recently: how strong AI is changing my work as an ObjectScript developer right now! Yeah, yeah, I know I'm not a very fast, it should have happened 2-3 years ago... But it's only now that I've realized that I am not writing the code all my workday like before.

Hospital readmissions cost the US healthcare system $17 billion annually. A significant portion of those readmissions are preventable — but only if the right information reaches the right clinician at the right moment: at discharge.

That's the problem I set out to solve with Smart Discharge Navigator, my entry for the InterSystems Programming Contest: AI Agents for FHIR 2026.