DjuraEarthquake engineering / Software
Seismic risk, from research to practice
I am an Assistant Professor at IUSS Pavia working on seismic design, loss and risk assessment, currently on SONATA, a project on the earthquake safety of non-structural elements. I co-founded Djura, which builds cloud tools for catastrophe risk engineering, and I lead development of the Built Environment Data platforms for experimental test data.
01 / Services
Data platforms and APIs
Web platforms, public APIs and databases for engineering and research data, from design to deployment.
Built Environment Data: 49,000+ data and metadata requests from 1,300+ users.Scientific Python packages
Engineering and research methods turned into documented, installable code.
djura: 10,000+ downloads on PyPI.Optimisation and analytics
Algorithms and data analytics for production and operations problems.
Cutting-pattern optimisation for a window and door profile manufacturer.
02 / Selected projects
Oct 2025 – Sept 2028
SONATA: Sensor-driven statistical tools to evaluate risks and manage safety in the built environment
Assistant Professor (RTDA)
Earthquake safety of Italy’s built environment with a focus on non-structural elements: shake-table testing, sensors, AI and image recognition to assess seismic damage, derive fragility functions and build a prototype risk-management system.
Show detailsHide details
Contribution
- Designed the prototype system architecture for the risk-management system.
- Co-authored the work on shake-table prototype testing, sensing and image-based detection of non-structural elements.
- Building the data pipeline linking sensor and camera outputs to damage and fragility analysis.
Outputs
- Prototype System Architecture for SONATA: Sensor-Driven Statistical Tools to Evaluate and Manage Safety in the Built Environment (2026)
- Shake-table prototype testing for sensor-based seismic damage assessment of non-structural elements (2026)
- Advancing sensing solutions for data-driven seismic damage assessment of non-structural elements: current challenges and future directions (2026)
- Towards Automated Image-Based Detection of Non-Structural Elements (2026)
Software developer
Djura platform
Cloud platform giving programmatic access to structural, seismic, risk and loss data.
7 engineering tools in production.
Show detailsHide details
Challenge
Methods for record selection, loss and risk assessment lived in research code that practising engineers could not easily run.
Contribution
- Designed, built and deployed the platform on AWS, with a Nuxt front end, Flask API and Supabase database.
- Implemented the loss function, record selection, EDP-IM, hazard consistency and IM conversion tools.
- Built the automatic report generation.
More outcomes
- 10,000+ downloads of the djura Python package on PyPI.
Outputs
- Conversion of Seismic Fragility and Vulnerability Models to Alternative Intensity Measures for Regional Risk Analysis (2026)
- Djura ground motion record selector: a software solution for earthquake engineering (2025)
- Bridging the gap: a software tool for efficient ground motion record selection in earthquake engineering (2025)
Lead software developer
Built Environment Data: Experiments
Web platform and API for sharing experimental test data from large-scale campaigns, used to validate and calibrate numerical models for vulnerability assessment.
49,000+ data and metadata requests from 1,300+ users.
Show detailsHide details
Challenge
Data from large experimental campaigns was spread across laboratories and formats, hard to find and reuse for calibrating numerical models.
Contribution
- Designed and built the whole platform: back end, front end and the documented public API, with OpenAPI docs and cURL examples.
- Containerised the service with Docker behind Nginx, with Redis caching, and deployed it on Eucentre infrastructure.
- Wrote the paper describing the platform, published in Earthquake Engineering & Structural Dynamics.
More outcomes
- 89 test campaigns from 22 laboratories, 160+ GB of data.
- 100% FAIR score.
- In production since 2024; migrated from MongoDB to PostgreSQL with no data loss.
03 / Software
- djuraPython toolkit for record selection, hazard consistency, vulnerability and loss modelling.
- Integrated Performance-Based Seismic Design (IPBSD)Design framework that limits expected annual loss (EAL) while targeting a mean annual frequency of collapse (MAFC).
- xgb-rhomutMachine-learning models for nonlinear and collapse response of single-degree-of-freedom systems.
04 / Recent publications
2026
Conversion of Seismic Fragility and Vulnerability Models to Alternative Intensity Measures for Regional Risk AnalysisEarthquake Spectra
2025
Impact of modeling decisions on seismic loss and fragility assessment of steel buildingsEarthquake Spectra
2025
Correlation models for next‐generation amplitude and cumulative intensity measures using artificial neural networksEarthquake Spectra
05 / Working with
- IUSS Pavia
- EUCENTRE
- INGV
- EPOS
- Torrefuerte