Founder @ NeuroNautix & Metadatapp | Scientist | FAIR (meta)data | Open Science Advocate
I am a (neuro)scientist dedicated to bridging the gap between preclinical research and digital interoperability. I founded NeuroNautix in 2025 and Metadatapp in 2026 to improve how we manage metadata and to accelerate ethical standards in animal research. Here is my personal website and blog.
I pioneered the FAIRRR framework — integrating FAIR data principles with the 3Rs (Replace, Reduce, Refine). My goal is to ensure that preclinical research is as ethical as it is reproducible. Read the five-part FAIRRR series on my blog.
- Metadatapp: An API-first platform for structured preclinical metadata management — open-source core at metadatapp-oss.
- NeuroNautix: Specialized consultancy in behavioral experiments, home-cage monitoring, and FAIR metadata (GitHub).
- Ontology Engineering: Lead on HCMO (Home Cage Monitoring Ontology, v0.4.1 on BioPortal) and MBO (Mouse Behavior Ontology).
- The Behaviour Forum: Administrator and moderator for a global community of behavioral scientists.
Ontologies & semantic curation
| Project | What it does |
|---|---|
| HCMO | Home-cage monitoring ontology built from the COST TEATIME HCM definition: modular Turtle sources, SHACL shapes, competency queries, versioned releases with Zenodo DOIs. |
| hcmo-kgqa-lab | Knowledge-graph question answering over HCMO (term retrieval → SPARQL → grounded answers). |
| HCM Ontology Mapper | Web app mapping home-cage monitoring systems to HCMO terms, with JSON-LD/CSV export. |
Trustworthy AI for science
| Project | What it does |
|---|---|
| The Ethical Debt | AI-assisted computational review of 1,300+ sources on data welfare in preclinical research, with claim-level TRUST scores. |
| Open Review Atlas | Proof-of-concept archive of AI-generated reviews with claim-level trust assessments. |
| FAIR.md · TRUST.md | Experimental machine- and human-readable conventions declaring a project's FAIRness and the evidence behind its claims. |
| Murmuration | Supervised LangGraph multi-agent literature review with reviewer checkpoints and an audit trail. |
| Research-Deposition-MCP | MCP server where agents prepare Zenodo deposits, executed only after signed human approval. |
| PICO | Pi Coding Orchestrator: deterministic Claude/Codex routing, provenance and isolated workers. |
FAIR research infrastructure
| Project | What it does |
|---|---|
| OpenResearchWorkspace | Portable RO-Crate research workspaces. |
| RELAY | Experimental profile-driven exchange format and validator for lab-study metadata (ISA, PROV-O, RO-Crate). |
| Data2Agent | Evidence-preserving access to scientific datasets through MCP, with FAIR assessment. |
| FAIR-VCG-mentor | LLM-assisted tool for FAIR preclinical data and Virtual Control Groups (3Rs). |
| nam2evidence | Proof-of-concept packaging of NAM-derived nonclinical data into reviewable evidence. |
| orcid-playground | Experimental open-source ORCID tools. |
Open hardware
| Project | What it does |
|---|---|
| Open-BEATBox | Open behavioural apparatus: hardware (CAD, KiCad), Qt/MicroPython software and assembly documentation. |
| Domain | Specialization |
|---|---|
| Bio-Behavioral | Home Cage Monitoring (HCM), ECG/HRV analysis, Skin-Nerve recordings, Social Fear Conditioning |
| Data Science | Python (Pandas, SciPy, Seaborn, statsmodels, DABEST), Jupyter, Folium |
| Semantics | RDF, OWL, SHACL, SPARQL, JSON-LD, Schema.org, RO-Crate, Ontology Mapping |
| AI & Agents | LLM-assisted curation, LangGraph, MCP servers, human-in-the-loop review |
| Development | JavaScript/TypeScript, API Design, HTML/CSS |
- DataCare is Animal Welfare: why FAIR metadata is the next 3R infrastructure
- FAIRRR 1/5 — The Metadata Gap Is an Ethics Gap
- FAIRRR 3/5 — Inside MAPP: API-First Metadata for Reproducible Research
- FAIRRR 4/5 — Case Study: Home Cage Monitoring with MAPP + Ontology Mapping
"Pushing for a description over an interpretation of data."
