Python • SAS • SQL | CDISC SDTM/ADaM • FHIR • OMOP CDM

I am a biochemist specializing in Clinical Data Engineering and Statistical Programming. I build reproducible and auditable pipelines across CDISC, FHIR, and OMOP CDM, combining biomedical domain knowledge with Python, SAS, SQL, data quality controls, and clinical terminology governance.

My portfolio follows clinical data from synthetic raw EDC or FHIR inputs to analysis- and RWE-oriented structures. It emphasizes traceability, deterministic processing, explicit validation boundaries, and human-reviewed semantic mapping. The projects are portfolio/reference implementations and do not claim regulatory validation or clinical production use.

I am seeking junior opportunities in CROs, pharmaceutical companies, RWE, and healthcare IT, in Portugal, across Europe, or remotely.

• Languages: Python, SAS, SQL
• Data Engineering: DuckDB, Pandas, ETL Pipelines, RAG (Retrieval-Augmented Generation)
• Clinical Standards: OMOP CDM v5.4, CDISC (SDTM/ADaM), FHIR, OHDSI Ecosystem
• AI & ML: Ollama, ChromaDB, Sentence-Transformers
• Software Engineering: Flask, PostgreSQL, Pytest, Git, GitHub Actions, CI/CD, Ruff, Data Governance

My portfolio forms a coherent narrative covering the entire clinical data lifecycle, from Raw Electronic Data Capture (EDC) to Real-World Evidence (RWE) ready formats:

A production-oriented reference framework that transforms synthetic FHIR JSON bundles into OMOP CDM v5.4.

• Combines deterministic terminology mapping with RAG and local LLM proposals.
• Implements fail-closed controls, blinded human review, provenance, and DQD checks.
• Includes CI, an extensive automated test suite, versioned benchmarks, and a reproducible release process.

A clinical data integration reference pipeline with production-oriented controls.

• Maps synthetic CDISC SDTM datasets into OMOP CDM v5.4 with record-level lineage.
• Uses deterministic and LLM-assisted terminology proposals behind a human approval gate.

An educational clinical programming pipeline built with synthetic study data.

• Transforms raw EDC-style inputs into CDISC-inspired SDTM and ADaM datasets.
• Demonstrates defensive SAS programming, QC, TLFs, and a structural Define-XML prototype without claiming submission readiness.

flowchart LR
    A["Raw EDC Data"] --> B["Clinical-data-to-CDISC"]
    B --> C["CDISC SDTM/ADaM"]
    C --> D["CDISC-to-OMOP"]
    D --> E["OMOP CDM v5.4"]
    F["FHIR JSON Bundles"] --> G["FHIR-to-OMOP"]
    G --> E
    E --> H["Real-World Evidence"]

Financial literacy is a personal interest of mine. As a complementary full-stack project, I built Amealha, a personal finance platform for tracking income and expenses, managing accounts, sharing household costs, and exploring financial scenarios. It demonstrates my broader software engineering practice across Flask, PostgreSQL, authentication, privacy, automated testing, deployment, and production monitoring.

Open to Junior Clinical Data Engineer, Statistical Programmer, Clinical Data Programmer, and RWE opportunities.