Computational systems scientist · Graph intelligence · Network science · Natural capital
I build graph-based intelligence systems that reveal structural dynamics, emergent risks, and hidden dependencies invisible to aggregate metrics. My work spans knowledge graph architecture, ecological network modeling, gene regulatory networks, blue carbon verification, and natural capital assessment, across biological, ecological, and decentralized systems.
- Knowledge graphs at scale: production systems with millions of nodes and relationships, encoded with domain semantics and queried via Cypher and GraphRAG
- Ecological network intelligence: food web reconstruction, k-core fragility analysis, cascade modeling, and biodiversity risk quantification across thousands of species and verified against hundreds of empirical food webs
- Natural capital baselines: investor-grade geospatial assessments integrating multi-sensor remote sensing (Sentinel-2, Landsat, SAR), CatBoost carbon stock modeling (R²=0.926), and Monte Carlo uncertainty at national scale (276K+ ha, 10K+ planning units)
- Nature-finance translation: Bio-VaR (Biodiversity Value-at-Risk), bond-style MPA ratings (AAA–CCC), ecosystem service valuation frameworks, and TNFD/ISSB-aligned biodiversity risk metrics for insurance and capital markets
- Blue carbon verification: VM0033/VCS-aligned pipeline with multi-GHG accounting via molecular markers (mcrA/pmoA), blue carbon site intelligence for community-led projects across Colombia and Ecuador (440K+ ha)
- Gene regulatory networks: GRN inference and analysis from single-cell data, community graph pipelines, and population structure visualization
- Agentic graph pipelines: multi-agent systems with LangGraph and GraphRAG for graph-grounded reasoning and discovery
Graph & networks: Neo4j · NetworkX · igraph · ggraph · tidygraph · LangGraph · GraphRAG Languages: Python · R · JavaScript (D3) · Julia · Mathematica Geospatial: GDAL · Sentinel-2/SAR · GEE · CatBoost · Monte Carlo simulation ML / embeddings: Protein language models (ESM-2) · Genomic language models (DNABERT-2) · scikit-learn Visualization: ggraph · visNetwork · D3 · Sankey diagrams
Conservation finance: $26B natural capital baseline (Gulf of California, 10,913 planning units), biodiversity credit frameworks, parametric insurance, and ecosystem service bonds for Marine Protected Areas
Blue carbon: Degradation cascade modeling (dNDVI/dNBR/dRVI) validated against world's first VCS-verified mangrove project; methane liability verification addressing the 27.5% CH₄ offset gap documented in Nature Geoscience (Qin et al. 2025)
Biodiversity risk for finance: Structural fragility layer complementing Swiss Re BES Index; network tipping point detection for reinsurance and TNFD disclosure
Open science infrastructure: Dependency graphs, transitive impact scoring, and network-based risk quantification for open-source ecosystems and public goods funding
Building open, verifiable graph infrastructure for decentralized ecosystems, dependency graphs, transitive impact scoring, and network-based risk quantification for public goods funding.