Software and data engineering for science & healthcare.
A data and software house. We build the toolchain, the OMOP CDM 5.4 and HL7 FHIR R4 data pipelines and the clinical software that research groups and healthcare providers use to run observational studies - and we sign what we hand over so you can verify it.
A data software house
Sigil builds the software, data and AI platform that research groups and healthcare providers use to turn clinical data into evidence. Foundational tooling, data engineering, clinical software and secure delivery - each stage reproducible, auditable and signed.
Eight tool families, listed below. The estate is private, so you cannot count it from outside; under NDA we walk you through the repositories behind whichever family matters to you. Everything we build is handed back as repositories you own and can run yourself.
One toolchain, end to end
A cohesive platform that carries a study from raw data to a signed deliverable, grouped into the tool families our teams and partners build on. The repositories are private; we walk you through the relevant ones under NDA.
Toolchain & GitOps
Govern the entire software estate as declarative, auditable state across every org and forge.
Data Engineering · OMOP and FHIR
Harmonise EHRs, claims and registries into OMOP CDM 5.4 and HL7 FHIR R4 - without moving the data.
RWE Study Packages
Entire observational studies authored declaratively and executed the same way against any site.
Analytics Platform
OHDSI ATLAS, WebAPI and Achilles deployed as one project. A cohort-authoring assistant is in development.
Dashboards & Visualisation
Population and single-patient clinical views from the same spine - fail-closed, FHIR-native.
Vocabulary & Standards
Source-code to standard-concept mapping with national and local conformance, as overlays.
Signing & Compliance
PAdES PDF signatures, each logged to the public Sigstore Rekor transparency log. Every signed PDF carries a QR to its own log entry, and the verifier is open to anyone.
Agent Fleet & Automation
A governed fleet of headless agents running as CI gates with structured verdicts. A human authorises every merge.
Ways teams bring us in
Four ways teams bring us in: stand up a platform, run a study, embed engineers, or advise on standards. Everything we build is handed over as repositories you own.
Build your platform
We stand up the foundational toolchain, data layer and GitOps model your teams build on, then hand you the keys.
Run a study, end to end
Fixed-scope engagement. We take the protocol, build the pipeline and the analysis, and hand back the signed report plus every repository used to produce it.
Embed with your team
Our engineers work inside your repositories, on your branches, to your review rules.
Advisory & standards
Architecture, security and interoperability guidance across OMOP CDM, HL7 FHIR, OHDSI and verifiable signing.
Automated, self-testing and secure - by design
Deep requirements understanding
Fast, open iteration cycles converge requirements early and keep delivery transparent.
Automated by design
Every pipeline is declarative and re-runnable from a clean checkout. Human authorisation gates are explicit, not implicit.
Self-testing by design
Tested against synthetic datasets in sandboxed environments before anything touches production data.
Secure by design
Least-privilege access and secrets handled as references, never values - from the first commit.
Agent-assisted delivery
Headless agents run as CI gates and emit structured verdicts. A human authorises every merge.
Handover, not lock-in
You receive the repositories, the pipelines and the deployment manifests. Nothing runs only on our infrastructure.
Research by our partners
Work by our research partner MTG Research and collaborators at the University of Porto. Sigil is not an author on these papers. Two concern generative models for molecule design; one concerns federated learning across paediatric care sites.
Novel Molecule design with POWGAN, a Policy-Optimized Wasserstein Generative Adversarial Networks
A policy-optimized Wasserstein GAN for de novo molecule design, extending reinforcement-learning-guided generative modelling. Preprint, not yet peer-reviewed. University of Porto.
MedGAN: optimized generative adversarial network with graph convolutional networks for novel molecule design
A generative adversarial network coupled with graph convolutional networks for de novo molecule design, demonstrating deep generative modelling for drug discovery.
Secur-e-Health Project: Towards Federated Learning for Smart Pediatric Care
A federated-learning architecture for smart paediatric care, training models across institutions while the data never leaves its source. MTG Research and Development Lab, with CINTESIS@RISE and GECAD.
Six people
A physician-founder, three software engineers, an operations lead and a strategic advisor.
Anna Cederlund
Vasco Colaço
Where we stand in the OMOP ecosystem
Sigil is not an EHDEN-certified SME, not listed on the OHDSI Portugal national node, and not a DARWIN EU data partner. Those are held by organisations, and none of them is held by this one. What is on the record sits next door, and we would rather point at it precisely than borrow it.
Our research partner is on the register
MTG Research & Development Lab is a trained and certified EHDEN SME and is listed as an SME of the Portuguese OHDSI national node. That work is MTG's, not Sigil's.
Our founder built that practice
Tiago Taveira-Gomes founded MTG Research & Development Lab and was interviewed by EHDEN in 2023 on harmonising Portuguese data partners to the OMOP CDM. He brings that experience to Sigil; the certification stayed with MTG.
We build on the OHDSI stack
ATLAS, WebAPI and Achilles deployed as one project, with data landed in OMOP CDM 5.4, and ETL targeting the EHDEN and DARWIN EU network models. That describes our tools, not standing in those networks.
Partners
Organisations we build with. MTG Research is our research partner and a certified EHDEN SME - the papers above are theirs, not ours.
Get in touch
We work with research institutions, healthcare providers and industry. Tell us the data sources, the standard you need to land in, and the deadline.
DMCC Business Centre, Dubai, 4076, AE
Tell us what you need built
A single study or a full platform. Tell us the data sources, the standard you need to land in, and the deadline.