Edimah SYNESIUS SONGO
Applied Mathematician · Reliable, Mathematically Grounded AI Systems · Bayesian Inference · EU AI Act
I’m an applied mathematician. The constant in my work is mathematical structure — Bayesian and frequentist inference, optimisation for ML, kernel methods, deep learning, signal processing — understood deeply enough that the same ideas carry across domains. Energy, public health, and finance have been the surfaces; the mathematics is the invariant.
I build reliable, mathematically grounded AI systems for regulated environments: Bayesian calibration of physics models at EDF, variance-based sensitivity analysis for high-dimensional systems at IFP Energies nouvelles, and today independent work in statistical modelling, AI reliability, and EU AI Act readiness — alongside an interim Data Lead role rebuilding the end-to-end data chain of a public cancer-screening organisation. Reliability is a design constraint I engineer for, not a job title.
Everything on this site asks one question of the systems it studies: how would I know whether to trust this?
What I am working on:
- Statistical frameworks for AI conformity assessment (EU AI Act risk tiers)
- Data chain audit methodology for regulated public health systems — case study in progress