Researcher · Engineer · Dublin
Saheed Faremi.
Researcher of the brain. Engineer of the systems people rely on.
01 · About
A researcher who ships.
Two modes. My doctorate decodes recurrent patterns in brain signal towards real-time detection of brain disorder. My engineering practice ships the production systems that under-served users (geographically, economically, or computationally) rely on.
The doctoral side: EEG microstates are quasi-stable scalp topographies that segment continuous brain signal into a discrete temporal alphabet. Find the recurrent patterns and you have a candidate biomarker. The work asks whether deep generative models, variational autoencoders and Gaussian-mixture VAEs, can learn a microstate alphabet that's more stable across sessions and more behaviourally predictive than classical clustering. Target: a representation reliable enough to anchor real-time detection of disorder-relevant brain states.
Production work spans fintech (Curnance), HR (Etihuku), healthcare data (HIS Core, Predict-Dx), advisory tooling (Gatsheni Advisory, CFI Eswatini), AI-assisted proposal generation (Cleva, built for Gijima), and learning systems (Skills Hub, Moodle). Infrastructure for under-served users needs the same engineering rigour as infrastructure for everyone else. It tends to need more.
In 2022 I represented Eswatini at the UNESCO India-Africa Hackathon at Gautam Buddha University in Uttar Pradesh. Team Geeks_on_Fire (five countries, five people) won problem statement AGRI12: an AI-assisted voice contact centre that lets farmers without smartphones report issues by phone and receive guidance back in their language. Gold medals and a ₹3 lakh team prize.
Based in Dublin, Ireland. Travel for research. open to collaboration
Find the recurrent patterns and you have a candidate biomarker.
02 · Research
EEG microstates with deep generative models.
EEG microstates are quasi-stable scalp topographies, typically four to seven canonical classes, that segment continuous EEG signal into a discrete temporal alphabet. The classical approach uses modified k-means clustering over the global field power maxima. It works, but it depends on hard choices (the number of states, the reference electrode, the band-pass) and the resulting segmentation can be brittle across sessions.
This project asks whether a learned latent geometry, via a variational autoencoder, produces a microstate alphabet that is more interpretable, more stable across sessions, and more predictive of behaviour than the classical pipeline.
Approach
- VAE. Single Gaussian latent prior; learn a continuous embedding of topography frames; segment by latent-space clustering or by direct decoder reconstruction error.
- GMM-VAE. Gaussian-mixture latent prior with one component per microstate class, so the segmentation falls out of the latent prior structure rather than a post-hoc clustering step.
- Architecture search. Sweep over latent dim, regularisation, and decoder choices; compare reconstruction-vs-segmentation tradeoff curves across architectures.
Recent publications
- Integrating Convolutional Variational Autoencoders and the Gaussian Mixture Model for efficient manifold learning and clustering of spatially preserved EEG topographic maps
Saheed Faremi, Luca Longo
Brain Informatics (under review) ·2026
- Interpretable EEG Microstate Discovery via Variational Deep Embedding: A Systematic Architecture Search with Multi-Quadrant Evaluation
Saheed Faremi, Andrea Visentin, Luca Longo
XAI 2026 (Late-breaking work + Doctoral Consortium track), Fortaleza, Brazil. arXiv preprint. ·2026
- Explainable Disentangled Representation Learning of Recurring Brain Activation Patterns via Variational Autoencoders
Saheed Faremi
XAI World Conference 2025, Doctoral Proposals track ·2025
- Machine Learning Models for Identifying Factors Influencing and Predicting Malaria Among Children Under Five Years in Nigeria
Akinpelumi Saheed Faremi, Boluwaji Akinnuwesi, Elliot Mbunge, Petros M. Mashwama, Stephen Fashoto, Polite Zenzo Ncube, John Batani, Shamsudeen Ademola Sanni, Yinusa A. Faremi, Andile Metfula
IEEE ICTAS 2024 ·2024