Nikolai Ryzhikov

How SQL on FHIR makes healthcare analytics interoperable: ViewDefinition, the new SQLQuery and SQLView profiles, the run/export/materialize API, and a dbt-style ELT pipeline for FHIR to OMOP.

CodeableConcept through the producer/consumer lens: why one field holds multiple codes, what the producer must capture while they still can, and how the consumer should interpret the array — membership, subsumption, projection — without .first() biting them.

Give your coding agent a REPL inside its own runtime — and it will extend itself to fit you, your project, and the task.

How FHIR R5 additional bindings solve the limitations of the classic one-binding-per-element model with layered terminology constraints.

We design agent tools as if agents were junior developers. They're not. Give them raw protocols and they'll outperform our carefully crafted abstractions.

R6 moves immature resources out of core into incubator IGs. Here is how additional resources work, what changed, and what it means for implementers.

What residual categories mean in FHIR terminology, how they affect ValueSet authoring, and pitfalls to avoid when modelling clinical concepts.

FHIR's nested structure works for APIs but frustrates SQL analytics. SQL on FHIR uses a simple DSL to define flat views – portable across ETL/ELT runners, tested via shared suites. Nikolai Ryzhikov's DevDays session demos the workflow from Synthea data to Grafana dashboards.

Master unified FHIR async patterns for bulk exports and other long-running operations using Bulk Data Access and R5/R6: Prefer headers (bundle/redirect/manifest), polling, callbacks, long polling, plus HTTP examples and visual flows for Manifest, Bundle, Redirect, Cancel, and Callback workflows.

SQL on FHIR in PostgreSQL. Two types of SQL on FHIR View Runners - ETL (in-memory) & ELT (in-database) and demo the ELT implementation for PostgreSQL.

What a FHIR ViewDefinition is, how it flattens FHIR resources for SQL analytics, and how to use it in the SQL on FHIR specification.

Master Patient Index (MPI) and healthcare Master Data Management explained: how record linkage works, common architectures, and a worked example using FHIR and Aidbox.

Why FHIR terminology splits into authoring and runtime phases — code systems, value sets, expansion denormalization, and a FHIR server's role.

If you are interested in FHIR databases and analytics on FHIR and going to be not far from Baltimore September 29–30 — please join us at the FHIR Connectathon 19 and brand new Storage and Analytics track!

SQL on FHIR: what it is, why it matters for healthcare analytics, and how it lets you query FHIR resources with familiar SQL syntax.

Using FHIR for data persistence is becoming a popular topic of discussion in the community. Our CTO shares his thoughts on when and why you would use FHIR as the storage format for your healthcare data.

Our CTO provides a brief technical overview of the HL7 FHIR Standard, from analysis of HL7 FHIR healthcare informational data model to potential use cases using FHIR APIs and potential limitations you should be aware of when planning to develop using FHIR standard.

Many people are talking about microservies as the preferred approach to structure complex health information systems. But there are overlooked consequences to using microservices over monolithic architectures.

Making sure your healthcare data is secure is a top priority for any health information system. Here’s what you should think about when managing access control for your FHIR server.

The FHIR standard is becoming the de facto way to manage healthcare data. Our CTO explains why FHIR implementations with dynamic languages could provide valuable feedback to the FHIR standard and FHIR community.

Deciding between which healthcare data standard to use can be hard. We evaluated whether HL7 v2, v3 & CDA, or FHIR is the best choice for a laboratory integration project. We found FHIR is the best choice because the FHIR API perfectly compliments the custom API of the laboratory system.