I am an independent researcher bridging data engineering with an academic background in law. My work focuses on building the open-science, reproducible data infrastructure that applied canine ethology desperately requires to overcome its current replication crisis. By shifting the paradigm from fragmented, subjective narratives toward deterministic data pipelines, my objective is to replace dogmatic assertions with traceably verified, physical realities.

Core Philosophy: Data Over Dogma True scientific rigor requires building scalable infrastructure to test behavioral claims objectively, moving the industry away from institutional appeals to authority and toward transparent code.

An open-science, Python-based Extract, Transform, Load (ETL) pipeline engineered to serve as a strict digital gatekeeper for ethological data. It transforms unstructured, messy observations into machine-readable datasets viable for quantitative peer review.

• Structural Reliability Matrix: Utilizes a validation engine built on Python and Pydantic to instantly reject malformed, subjective inputs (e.g., rejecting qualitative tags like "angry" or clamping physiological metrics to valid veterinary baselines like 30–250 BPM).
• Deterministic Semantic Parsing: Integrates Google Gemini 2.5 Pro as a strict text parser. By clamping the model's temperature parameter to 0.0 and enforcing Structured JSON Outputs, it extracts data matching peer-reviewed operational definitions without the risk of LLM hallucinations.
• Global Interoperability: Rigorously maps internal variables to Darwin Core (DwC) metadata standards (e.g., dwc:individualID, dwc:measurementType, and dwc:basisOfRecord as 'HumanObservation') to enable seamless integration with international informatics platforms like GBIF and OBIS.

A global analytics dashboard engineered to track data regarding aversive training legislation and behavioral euthanasia outcomes across multiple international jurisdictions.

• Architecture: Deployed via a Hybrid Framework combining a polished presentation layer on GitHub Pages with a live Python server hosted on Streamlit Community Cloud, integrated with a Firebase backend.
• Legislative Baselines: Audits and visualizes legal frameworks, including localized benchmarks such as São Paulo Municipal Law No. 18.314/2025, which regulates punitive handling tools.

• Languages & Core Development: Python, Structured JSON, HTML5/CSS3
• Data Engineering & Validation: Pydantic, ETL Pipeline Design
• AI & Cloud Infrastructure: Google Gemini 2.5 Pro API (Deterministic Optimization), Firebase, Streamlit Community Cloud, GitHub Pages
• Metadata & Informatics Standards: Darwin Core (DwC) Schema Mapping

To maintain complete transparency and integration within the peer-reviewed ecosystem, my contributions are traceably linked across platforms:

• ORCID iD: 0000-0002-1825-0097 (Example format—replace with your actual iD)
• Digital Object Identifiers (DOIs): Automated release pipeline linking GitHub tags to Zenodo to push verified open-source software data straight to my official scientific record.
• Manuscript Compliance: Embedded YAML metadata schemas matching JOSS (Journal of Open Source Software) automated compilation criteria.

If you are a researcher, data scientist, or epidemiologist interested in expanding reproducible, open-source frameworks for animal welfare, feel free to open an issue or reach out:

• Global Dashboard: thetransparencyproject.me
• Local Case Study Blog: rotapetplanaltopaulista.com (Specialized Portuguese landing page for São Paulo legislative tracking)

-->