Geospatial Data Engineer · Python · AWS · GIS · AI Workflows
Geosciences-trained technical operator. Remote-first. Colombia 🇨🇴
I build data systems that turn messy scientific and operational data into usable decisions — from underground geological surveys to cloud-scale document automation pipelines.
Career arc: Geoscience fieldwork → GIS development → AWS data engineering → Geoscience software & AI-augmented workflows
Currently: Geoscience Software Specialist at Imaged Reality (UK, remote) — Python API extensions, supervised ML classification pipelines, and enterprise client delivery on the Stratbox platform for major oil & gas operators.
Languages Python (Pandas, GeoPandas, PySpark, scikit-learn, Dash, Plotly) · SQL
Cloud AWS — Lambda · Glue · Step Functions · Textract · Athena · Redshift · CDK
GIS ArcGIS Online/Pro · ArcGIS API for JS · Survey123 · QGIS · Leapfrog Geo
ML / Data Random Forest · supervised classification · ETL design · Power BI
AI & tools Claude Code · MCP servers · Obsidian · agent-assisted workflows (daily)
Dev Git · Docker · SonarQube · Postman · DBeaver
| Project | Problem solved | Stack |
|---|---|---|
| AB InBev Invoice Pipeline | PDF extraction across 30–40 supplier formats — eliminated manual matching at scale | Textract · Lambda · Step Functions |
| DIMAR El Niño Dashboard | Oceanographic monitoring for Pacific expedition; COVID geospatial tracking for DIMAR | Python/Dash · ArcGIS Online · GeoPandas |
| Casaideas SAP Pipeline | Retail master data from SAP to cloud with traceable lineage | Glue · PySpark · S3 · CDK |
| Stratbox Python API | Depth-referenced extraction method + Random Forest facies classifier | scikit-learn · Python · AWS |
| Coscuez 3D Geological Model | First 3D sedimentary emerald resource model accepted by Colombian national authorities | Leapfrog Geo · DATAMINE |
Active
| Repo | What it is | Stack |
|---|---|---|
web-portfolio |
This portfolio site — projects, blog, engineering notes | Astro · TypeScript |
app-vienos-dashboard |
DIMAR oceanographic El Niño monitoring dashboard (Colombia Pacific) | Python · Dash · ArcGIS |
app-ritual-qr |
Duplex fold-and-backlight QR puzzle generator with QA + printer calibration | Python |
app-housing-simulator |
Rent-vs-buy housing scenario simulator with ROI metrics | Python · Streamlit |
skill-research-first |
Claude Code skill that enforces a research-first workflow before generating code | Claude Code · MCP |
Experimental / WIP
| Repo | What it is | Stack |
|---|---|---|
app-multi-bingo |
Multi-card bingo tracker for parallel games | Python |
lab-las-explorer |
Notebooks exploring well log .las files |
Jupyter · Python |
Reference / archived
| Repo | What it is | Stack |
|---|---|---|
lab-python-snippets |
Personal Python reference notebooks (collections, JSON, pickle, parquet) | Jupyter |
lab-computational-methods-coursework |
Computational methods coursework — Universidad de los Andes (2015) — archived | C |
📄 Automating Core Data Integration: Insights from the Volve Field Case Study
Borya, A., Henao, K., Ruiz-Graham, C. — ADIPEC 2025 — SPE OnePetro D041S138R004