Skills
What I’m best at
I deliver end-to-end modelling: define the decision and metrics, build the data foundation, develop features and models, validate rigorously, and ship reproducible outputs. Strength: statistical judgement and uncertainty-aware inference (especially hierarchical/Bayesian), useful when data are noisy, sparse, biased, or structured in space/time.
What I deliver
- Full-cycle modelling delivery: define objectives and success metrics, design the data foundation, engineer features, train models, validate rigorously, and deliver reproducible outputs that support decisions.
- Model selection with statistical judgement: select model families aligned to the data-generating process, including structured dependence across space and time.
- Bayesian hierarchical inference: partial pooling and principled uncertainty propagation for robust estimates and uncertainty-aware decision-making under sparse, noisy, or biased data.
- Experimentation and causal reasoning: A/B testing fundamentals, power and effect-size framing, and clear treatment of confounding, selection bias, and the limits of identification.
- Reliable evaluation and communication: leakage checks, calibration awareness, error slicing, robustness and stress testing, and transparent reporting of assumptions, tradeoffs, and limitations.
Core stack
Tools
- Python: pandas, NumPy, scikit-learn; pipelines; visualisation (matplotlib/seaborn/plotly)
- SQL (PostgreSQL): joins, CTEs, window functions; analytics transformations
- Software engineering: Git/GitHub; modular code; testing/validation; reproducible environments (conda/venv)
- LLM engineering: structured prompting, tool calling, RAG foundations, agent orchestration (LangGraph; Python), API integration
Core methods
- Bayesian & hierarchical modelling: partial pooling, uncertainty quantification and propagation
- Machine learning: supervised/unsupervised, model selection, tuning, gradient boosting (incl. XGBoost)
- Experimentation & causal: A/B testing, power/effect size, confounding, selection bias
- Evaluation discipline: leakage, calibration, error analysis, robustness
Academia transferability
- First-principles problem solving: break down open-ended questions, formalise hypotheses, and select methods aligned to the data-generating process.
- Study and measurement design: design sampling and measurement protocols, define data quality standards, and handle bias, missingness, and uncertainty at the source.
- Full-cycle project delivery: scope work, set milestones, manage tradeoffs, and deliver high-quality outcomes under real constraints.
- High-standard technical communication: publish peer-reviewed work and produce clear, auditable narratives with explicit assumptions, evidence, and limitations.
- Cross-functional collaboration and mentorship: work across disciplines, mentor and upskill collaborators, and drive alignment through proposal-style writing and resourcing justification (grants/funding).
Technical depth
Bayesian, hierarchical & spatiotemporal modelling
- Generalised linear models (GLMs) and generalised additive models (GAMs) for nonlinear effects
- Threshold and segmented regression for decision-point inference
- Hierarchical and mixed-effects modelling; partial pooling
- Bayesian inference with uncertainty quantification and propagation; priors as explicit assumptions
- Spatiotemporal modelling: structured dependence, forecasting logic, species distribution modelling (SDMs)
- Observation vs process modelling: detection–abundance separation; N-mixture models
- Integrated Population Models (IPMs)
LLMs, prompt engineering & agents
- Prompting for structured outputs; reliability patterns (prompt scaffolds, self-checks, evaluation loops)
- Tool calling and retrieval patterns; schema/contract design for model outputs
- Agent workflows: planning/acting loops, orchestration, retries, human-in-the-loop checkpoints
- LangGraph concepts: state, control flow, tracing/debugging
- OpenAI API integration patterns for prompt-driven applications
Machine learning (classical)
- Supervised learning: linear/logistic regression, tree-based models, random forests, gradient boosting (incl. XGBoost), SVM, k-NN
- Unsupervised learning: PCA, clustering (K-Means), anomaly detection
- scikit-learn Pipelines; hyperparameter tuning; model comparison and baselines
Deep learning
- Neural networks for classification and regression
- Training fundamentals: loss functions, optimisers, regularisation, monitoring and early stopping
- TensorFlow / Keras: model definition, training, evaluation
Evaluation, interpretability & reporting
- Validation design: train/val/test, cross-validation, temporal/blocked splits where appropriate
- Evaluation discipline: leakage checks, calibration awareness, error analysis and slicing, robustness/stress testing
- Interpretability: feature importance, partial dependence, SHAP-style global/local explanations
- Decision-ready reporting: assumptions, limitations, tradeoffs, and clear recommendations
Experimentation & causal inference
- A/B testing fundamentals: hypotheses, metrics, power and effect size
- Causal inference basics: confounding, selection bias, counterfactual framing, limits of identification
- Practical decision-making under uncertainty: interpreting results and communicating tradeoffs
Data engineering & integration
- Data ingestion and transformation: structured files, schema discipline, reliable I/O
- SQL-centric data work: joins across complex relational datasets, analytics transformations
- API integration patterns: extracting, normalising, and joining external data sources
Software engineering & reproducibility
- Git/GitHub workflows: branching, pull requests, code review, merge discipline
- Maintainable codebases: modular architecture, clean interfaces, reusable components, pipeline-style structure
- Quality controls: input validation, assertions, unit tests (pytest patterns), docstrings, type hints where useful
- Reproducibility: environment management (conda/venv), deterministic runs, versioned artefacts, methods-first documentation
NLP, recommenders & text features
- Text preprocessing and inspection: normalisation, tokenisation, feature auditing
- Vectorisation: bag-of-words and TF-IDF; baseline classifiers (Naive Bayes)
- Recommender foundations: similarity metrics, collaborative filtering, constraint-aware framing
Geospatial & remote sensing
- Raster/vector workflows; spatial joins; geoprocessing pipelines
- Spatial feature engineering; landscape/canopy metrics
- Scalable spatial processing
Visualisation & lightweight apps
- Visualisation: matplotlib, seaborn, plotly; ggplot2
- Lightweight apps: Streamlit (Python), Shiny (R)
Education & Training
A detailed, certificate-linked list of formal education and courses is maintained here: Education & Training.