API Artefacto 360 DAGMA
Field-capture backend for DAGMA-360: a FastAPI service backed by real-time Firestore, reconnaissance photos stored in Amazon S3, GPS coordinate capture, and metrics exposed via Prometheus.
- Python
- FastAPI
- Firebase
- AWS S3
Two worlds converging
AI · Data · Software
- Industrial RAG
- Advanced data science (NLP, spatiotemporal models)
- Data engineering (GCP, ETL, Data Warehouse)
- Geointelligence & GIS
- Frontend development (React)
- Data governance
- Web scraping
- Software consulting
Civil Engineering
- BIM 3D/4D (European standards)
- Hydraulic design & simulation
- Structural design (NSR-10)
- Territorial planning · Construction 4.0
- Public infrastructure project management
Featured experience
View all →DAGMA (City of Santiago de Cali)IT/AI Solutions Consultant
2026- Led the design of the DAGMA-360 architecture to integrate scattered environmental data (air quality, biodiversity, early-warning alerts) into a single territorial monitoring platform.
- Developed the SIMAP geospatial intelligence module to track vegetation cover and urban biodiversity from satellite imagery.
- Designed a local AI infrastructure (self-hosted models) to reduce dependency on external APIs and safeguard the municipality's environmental data sovereignty.
Stack:GCPGISPythonRAG
Cali Secretariat of Security and JusticeSenior Data Scientist
2024- Built, together with the Security Observatory team, a Data Warehouse fully hosted on GCP (SILOS project) to consolidate crime, coexistence, and citizen-perception data sources.
- Designed spatiotemporal risk models to anticipate incident concentration by zone and time band, feeding directly into operational planning.
- Implemented a press and social media scraper with an NLP pipeline to measure security perception, feeding React dashboards used by decision-makers.
Stack:GCPPythonNLPReact
Cali Secretariat of GovernmentTechnical Data Advisor
2024- Technically led the What Works Cities certification process, through which Cali became the fourth city in Colombia to earn it, validating the use of data and evidence in public management.
- Designed indicators and tracking dashboards for government programs, standardizing how the Secretariat reports progress with verifiable data.
Stack:Power BISQL
Naska DigitalBIM Specialist / Instructor
2024- Trained design teams in 3D/4D BIM workflows under European standards, from modeling through discipline coordination.
- Developed training materials and supported the adoption of Revit and Navisworks on active infrastructure projects.
Stack:RevitNavisworksBIM
Projects
View all →SSJ Data Warehouse
Backend for the Security Observatory's Data Warehouse (Cali's Secretariat of Security and Justice): a FastAPI API on PostgreSQL, with versioned migrations, Docker containers, and deployment on Railway.
- Python
- FastAPI
- PostgreSQL
- Docker
AutoML4RainEstimation
Rainfall estimation with AutoML and deep learning from weather data and satellite imagery, automating model selection, hyperparameter tuning, and feature engineering for climate forecasting and disaster risk management.
- Python
- AutoML
- Deep Learning
- Jupyter
From the blog
View blog →The NumPy reference sheet I actually keep open while working
Array creation, indexing, broadcasting, axis-aware aggregations, linear algebra, and random generation: a practical NumPy reference for real data work.
Read more →The Pandas Cheat Sheet I Actually Reach For
Selection, cleaning, grouping, merges, dates, and performance: the complete Pandas reference I keep open in every real data project.
Read more →Colombia's National Disaster Risk Management System, Explained
How Colombia's National Disaster Risk Management System works: its legal foundation, three pillars, and how the municipal level fits the national one.
Read more →