At the intersection of data science, mathematics, and research supporting informed decision-making, I’ve spent more than two decades building systems that turn complex data into actionable insights. It began unexpectedly in Afghanistan in 2002, where, with a background in electrical engineering, I found myself managing humanitarian programs across food security, water, and infrastructure initiatives. Working with limited information while making decisions that affected communities in Faryab province, I recognized a fundamental need: the ability to rapidly process vast amounts of data, identify patterns, and support time-sensitive decisions. That need has driven my work ever since.
Current work
I consult mainly for the IAEA, on data science in the broad sense:
Soil mid-infrared spectroscopy: research on using spectroscopy to guide the remediation of farmland contaminated by radionuclides, with the Joint FAO/IAEA Centre of Nuclear Techniques in Food and Agriculture.
Coffee plant water stress: data science support for an early warning system that combines stable isotope techniques and IoT sensors.
DSS4NAFA: developing capacity-building material for this decision support system for nuclear applications in food and agriculture.
MARIS: developing the data ingestion architecture of the marine radioactivity information system.
Training: delivering courses on data science and AI.
AI runs through all of these projects. I use it mainly to capture domain experts’ knowledge and turn it into tools that others can use, such as language-model access to the MARIS knowledge base. My focus remains on practical deployment in operational environments, adapting these technologies to real-world constraints rather than chasing novelty.
Earlier work
As a consultant with the IAEA since 2012, I’ve designed and developed decision support systems that bridge complex data with practical application:
IRMIS (International Radiation Monitoring Information System): developed the data science and visualization components of the first version, and advised on the technical design of the second, now used worldwide.
DSS4NAFA: designed and developed the data science components of the system.
With IOM, I developed LLM-assisted tagging of evaluation reports for evidence map production. That assignment is currently on hold.
Before that, I worked extensively with UNHCR and other UN bodies, specializing in refugee settlement mapping and geospatial frameworks for humanitarian response, from field operations to HQ-level coordination and interagency working groups.
About this blog
On this blog, I share what I learn from applying data science and AI to real-world problems. Consider it a workbench where practical experience meets ongoing curiosity. I also explain research papers I consider important, staying grounded in the mathematics that underpins this work.
For more details about my current consulting work, visit my LinkedIn profile.