Senior Data Scientist • Mathematician • based between the United Kingdom and Portugal
“Knowledge is knowing a tomato is a fruit; wisdom is not putting it in a fruit salad.”
— Miles Kington
I build production systems that turn messy data into decisions. Two decades across logistics, health, and engineering taught me the value of lean models, clean code, and reproducible pipelines. Lately I’ve been shipping NLP and statistical modelling that helps teams reason about text and time series in real time.
- Machine Learning
Supervised & unsupervised learning, anomaly detection, time-series forecasting, optimisation. - Graph & Network Analysis
Social/interaction networks, graph theory, dynamic metrics, community structure. - Big Data Analytics
Pattern discovery in marketing, logistics, and urban systems (structured & unstructured data). - Mathematical Modelling
Differential equations, statistical inference, numerical methods for complex systems. - Sustainability & Urban Systems
Energy optimisation, smart environments, traffic prediction.
- Programming — Python (typed, NumPy-first), SQL, R, TypeScript, Bash/Zsh, C, Fortran
- ML / Data — NumPy, Pandas, Polars, FireDucks; scikit-learn, XGBoost/LightGBM; PyTorch, TensorFlow; Statsmodels
Focus: time series, anomaly detection, GLMs/IRLS, robust statistics - Data Eng & Streaming — Apache Kafka, Flink, Spark, Databricks; Arrow/Parquet; Apache Iceberg (lakehouse)
- Cloud & Storage — AWS S3, DynamoDB; PostgreSQL, MySQL, SQLite; MongoDB, InfluxDB
- DevEx & CI/CD — Docker; GitHub Actions, Jenkins; Poetry; pre-commit (ruff, mypy, pytest-cov); semantic versioning
- Testing & Quality — pytest, coverage, property-based tests (hypothesis); static typing; security linting (bandit)
- Health Data Science — real-time analytics from wearables/sensors, personalised baselines, clinical interpretability
- Graph Theory & Social Networks — interaction graphs, diffusion/contagion models, community & role discovery
- Big Data & Marketing Analytics — uplift modelling, sequence-aware attribution, lifetime value with drift control
- Sustainability & Energy Systems — demand forecasting, optimisation under constraints, carbon-aware scheduling
- Smart Environments & Sensor Networks — multimodal fusion (RSSI + activations), localisation, reliability modelling
- Behavioural & Labour Economics — micro-behavioural patterns, incentive effects, heterogeneity and fairness
- Inequality & Sustainable Development — distributional metrics, policy simulation, causal and counterfactual analysis
Now: real-time anomaly detection; Bayesian filtering/HMMs for indoor localisation; robust regression & GLMs (IRLS); LLM-assisted reporting with audit trails; abx-next (modern A/B experimentation utilities).
-
abx-next — A/B experimentation utilities: CUPED/CUPAC hooks, triggered analysis, SRM guardrails, switchback helpers, and power simulations.
👉 repo -
genSurvPy — Survival-data generators (AFT/CPHM, censored data), reproducible simulations, and validation utilities.
👉 repo -
smart-todo-action — GitHub Action that extracts TODOs, groups by semantic labels/tags/metadata, and opens issues/changelogs.
👉 repo -
navier-stokes-solvers — CFD solvers for the 2D/3D Navier–Stokes equations (finite-difference & spectral variants), with buildable CLI targets and basic tests.
👉 repo -
heavytails — Utilities for heavy-tailed modelling and inference (tail index estimation, Pareto-like fits, EVT diagnostics).
👉 repo
- Introduction to Logic & Set Theory (First Semester, 15 weeks) — Logic (prop/FO), sets, induction, differential & integral calculus; notes + LaTeX.
- Linear Algebra (Second Semester, 15 weeks) — Vector spaces and linear maps; matrices and determinants; eigenvalues/eigenvectors, diagonalisation; orthogonality, projections, Gram–Schmidt; least squares; SVD and PCA; numerical stability & conditioning; applications to optimisation and data science.
Syllabus: link · Slides (Beamer): link - NLP & LLM mini-workshops — Prompt design, evals, lightweight retrieval, and report generation with structured → narrative transforms.
- Data Science Seminars — End-to-end ML pipelines, feature engineering for time series, evaluation under drift, MLOps (CI/CD, data/versioning), and reproducible research practices.
Slides: link · Notebooks: link - Sensors & Dashboards — IoT data ingestion (MQTT/Kafka), time-series storage (InfluxDB/Parquet), streaming analytics (Flink), and dashboards (Grafana/Plotly/Dash) with alerting & anomaly detection.
Slides: link · Demo repo: link - Applications of Matrices to Computational Graphics — Linear transforms in 2D/3D, homogeneous coordinates, rotations (Euler vs. quaternions), camera models & projections, shading basics; SVD/PCA for geometry processing.
Slides: link · Code samples: link
- Streaming analytics with Iceberg + Flink + DynamoDB — Architecture notes and example pipelines.
- Robust regression with IRLS — ψ-functions, influence diagnostics, and uncertainty reporting.
- Time-series anomaly detection — EWMA variants, adaptive σ, and change-point alerts for sensors.
- Interdisciplinary approach spanning computer science, mathematics, economics, and natural sciences.
- Practical projects in IoT, automation, and environmental monitoring (Raspberry Pi + sensors).
- Ongoing work in ML for time series, anomaly detection, and robust statistical modelling.
Thanks for visiting! I’m keen to partner with data enthusiasts, researchers, and product teams. Browse my projects or get in touch—happy to explore ideas and build useful things together.
