Project portfolio
A decade and a half of applied data work — climate, environment, public health, urban policy.
Capybara today builds open climate early-warning models. The portfolio below shows where the methods came from: malaria fieldwork in the Amazon, urban epidemiology in Bogotá, indigenous data governance, credit-scoring with alternative data, science-art collaborations. Each prior project contributed an idea or a method that survives in the current work.
Current flagship
Capybara Center / EWS
2024 — Ongoing
Founder · Lead analyst
An open-science climate early-warning project. Built and maintains a family of multi-hazard models (wildfire, flood, heatwave) running on Google Earth Engine, publishes JANUS-format situation reports for affected regions, runs an open GitHub repository for volunteer contributors, and operates the analytic portal at capybaracenter.blogspot.com.
Key outputs: situation reports for Western Europe, Mexico, Guatemala, El Salvador, Antioquia, Nuevo León, Colombia. JANUS framework (peer-reviewed 2025). Open code at MIT licence.
Past Experience of our team that nourishes this project
FAIR2Adapt Project — Stockholm Environment Institute
2025 · Research fellow
Designing diagnostic tools for stakeholder engagement and implementing FAIR (Findable, Accessible, Interoperable, Reusable) data principles in climate-adaptation projects. Leading participatory design for climate-risk dashboards and data-translation workshops across international consortia. This project inspired the adoption of the FDO framework to generate automated reports comparing model outputs with news media outlets.
Climate, environment & indigenous data
SIGETI — Information System for Indigenous Territory Governments
2022 · Lead designer · Gaia Amazonas Foundation
Conceived and implemented a participatory data platform for climate-change adaptation with indigenous communities in the Colombian Amazon. Combined remote sensing, field surveys, and local knowledge to deliver tools for territorial governance and climate resilience. Direct ancestor of the Capybara model family's emphasis on per-pixel granularity and community-validated outputs.
YOU:MATTER — Science-Art Experience · Marshmallow Laser Feast
2024 · Scientific consultant
Designed the scientific "bible" and narrative scaffolding for an immersive installation integrating climate and public-health data into large-scale interactive art. Collaborated with artists and developers to embed real data models into creative storytelling, reaching audiences that journal publications could not.
Public health & urban analytics
COVID-19 Public Policy & Urban Acupuncture — Bogotá
2020 — 2022 · Bogotá city + national collaborations
Co-designed mobile-phone-data-driven interventions, real-time dashboards, and quasi-experimental studies to inform Bogotá's outbreak response. Supported city and national agencies with spatial risk modeling, scenario planning, and impact assessment of micro-territorial interventions. The "urban acupuncture" framing (targeted interventions at leverage points) survives in how Capybara situation reports identify specific neighborhoods rather than department-wide alerts.
Urban & Epidemiological Analytics — Universidad del Rosario
2018 — 2024 · Research lead
Advanced methods in topological data analysis applied to epidemiology and urban ecology. Led projects on the spatio-temporal dynamics of infectious diseases (malaria, COVID-19) and urban ecological structure. Developed social-network analysis tools for evaluating political debate, consensus, and digital violence.
Malaria Heterogeneity & Quantitative Ethnography
2010 — 2022 · Multiple Amazon projects
Pioneering fieldwork on malaria risk in the Colombian Amazon, combining genomic, spatial, and ethnographic methods. Published peer-reviewed studies on disease transmission, demographic transitions, and the interface between human ecology and public health. This is the work that taught the methods that became the Capybara EWS spine — multi-source data fusion, hyperlocal risk, community ground-truthing.
Other applied data work
Credit Scoring & Social Impact Analytics — Agatha
2021 · Revenue & partnerships lead
Led the design of credit-scoring systems for underserved populations using alternative data sources (retail receipts, behavioral signals, mobility patterns). Supported advanced analytics across public health, mobility, and public safety. The methodology underwrites Capybara's current climate-aware credit work.
Migration, Conflict & Urbanization Analytics — Colombia
Multiple periods
Explored migration dynamics, urban growth, armed conflict, and anthropogenic change in Colombia using large-scale datasets and advanced network science. Spatial-temporal methods from this work inform how Capybara represents demographic exposure layers in the current models.
Coffee Roasting Analytics — Jaguar Coffee Co.
2018 — 2019 · Data-driven startup
Built one of the first "roasting curve analytics" platforms in Latin America, integrating sensor data and consumer feedback for quality improvement in specialty coffee. An early demonstration that the same analytic toolkit moves cleanly between domains as far apart as Amazon malaria and the coffee community. Most importantly, ethnographic work in this topic highlighted the need to generate information models to empower cooperatives of small-scale agricultural projects.
What all this work shares
A consistent commitment to: open data and reproducible methods, hyperlocal spatial-temporal granularity, multi-source fusion (satellite × administrative × behavioral × community), and translation from analysis into a form that decision-makers and communities can actually use. Capybara Center is the consolidation of these threads into a single, openly licensed, volunteer-supported project.
Capybara takes on partner engagements in climate early warning, spatial epidemiology, urban analytics, indigenous data governance, and FAIR climate adaptation. Engagements are project-based and shaped to the partner's specific context.
Start a conversation →
Capybara Center · capybaracenter.blogspot.com · Open code (MIT) · JANUS framework: DOI 10.21203/rs.3.rs-7580516/v1
Note: Only a member of this blog may post a comment.