Education & Training

• Focus: research-grade modelling for ecological systems under uncertainty and spatiotemporal structure.
• Skills & tools: Bayesian and frequentist inference; hierarchical modelling; predictive modelling; GIS/remote sensing; automated data pipelines; data cleaning; R; reproducible workflows.
• Data types: multi-source ecological, biological, climatic, physiological, and biogeochemical datasets (e.g., soil, foliage chemistry).
• Outcome: developed modelling frameworks to predict vulnerability to extreme events and identify high-risk habitats.

• Skills & tools: GIS; advanced statistics; applied statistical modelling; R; spatial analysis; workflow automation.
• Outcome: designed and executed analytical workflows for biodiversity monitoring and conservation planning.

• Skills & tools: mathematics, algebra, biostatistics, physics, introductory statistical programming, ecological modelling.
• Outcome: undergraduate research integrating environmental and ecological data.

• Statistical Rethinking: A Bayesian Course with Examples in R and Stan
• Statistical Rethinking 2023 — Online Course (Richard McElreath)
• Bayesian Methods for Ecology — Michael A. McCarthy
• Applied Hierarchical Modeling in Ecology
• Integrated Population Models
• Statistics in R workshop — Dr Murray Logan (AIMS)

• R for Data Science

• Outcome: practical, end-to-end R workflow from data handling to modelling and interpretation.
• Coverage:
  • Core R fundamentals (objects, vectors, indexing, functions) and interactive workflows.
  • Data handling with data frames; vectorised operations; tidy data principles.
  • Wrangling and visualisation with tidyverse and ggplot2.
  • Reproducible analysis with R Markdown; workflow discipline and code organisation.
  • Statistical foundations: introductory inference testing and interpretation.
  • Modelling coverage: linear models, GLMs/GLMMs, mixed-effects models, GAM/nonlinear frameworks, multivariate analyses (frequentist + Bayesian exposure).

(Click to view certificate)

• Outcome: strong programming foundation for automation, scripting, data handling, and maintainable code.
• Coverage:
  • Core Python: data types, control flow, functions, and standard debugging patterns.
  • Data structures and program design; writing clean, maintainable code.
  • Error handling and regex; iterators/generators; decorators (intermediate concepts).
  • OOP fundamentals: classes, inheritance, encapsulation.
  • File I/O and common modules (collections, datetime, os, math, random, re).
  • Working with CSV/PDF/image files; automation patterns and scripting utilities.
  • Web scraping with Requests + BeautifulSoup.
  • Jupyter + standalone .py development workflows; small end-to-end mini-projects.

(Click to view certificate)

• Outcome: end-to-end applied ML workflow: wrangling → modelling → evaluation.
• Coverage:
  • Data wrangling and numerical computing: pandas, NumPy.
  • Visualisation: Matplotlib, Seaborn, Plotly.
  • ML with scikit-learn:
    • Supervised: Linear/Logistic Regression, k-NN, Decision Trees, Random Forests, SVM, feed-forward neural networks (intro).
    • Unsupervised: K-Means, PCA.
  • Intro NLP: tokenisation, vectorisation (CountVectorizer/TF-IDF), Naive Bayes baselines.
  • Intro recommenders: similarity metrics and collaborative filtering.
  • Deep learning foundations with Keras (intro).
  • Evaluation workflows: cross-validation, train/test design, bias–variance reasoning.
  • Exposure to big-data concepts with Spark (intro).

(Click to view certificate)

• Outcome: strong conceptual + practical grounding in classical ML and core modern foundations.
• Coverage:
  • Supervised learning fundamentals; model evaluation and optimisation concepts.
  • Advanced learning algorithms; neural network foundations; decision trees and ensemble logic.
  • Unsupervised learning; anomaly detection; recommenders; introductory reinforcement learning.
  Module certificates

  • Module 1 — Supervised Machine Learning: Regression and Classification
    (Click to view certificate)

  • Module 2 — Advanced Learning Algorithms
    (Click to view certificate)

  • Module 3 — Unsupervised Learning, Recommenders, and Reinforcement Learning
    (Click to view certificate)

(Click to view certificate)

• Outcome: strong SQL foundation for analytics workflows in PostgreSQL.
• Coverage:
  • Core querying: SELECT, WHERE, filtering, ordering, pattern matching.
  • Aggregation: GROUP BY, HAVING.
  • Joins: INNER/LEFT/RIGHT/FULL/CROSS; multi-table workflows.
  • Schema fundamentals: tables, constraints, data types; practical loading/setup.
  • PostgreSQL tooling with pgAdmin; integrating SQL outputs with Python pipelines.

(Click to view certificate)

• Outcome: advanced analytical SQL patterns for complex relational datasets.
• Coverage:
  • Advanced joins, unions, and self-join patterns.
  • Subqueries and CTEs (including recursive CTEs) for multi-step logic.
  • Window functions: ROW_NUMBER, RANK, DENSE_RANK, LAG, LEAD, FIRST_VALUE, LAST_VALUE.
  • String/numeric/date functions; NULL-safe handling and conditional expressions.
  • Analytics patterns: de-duplication, conditional aggregation, segmentation, pivot-style summaries.
  • Views and modular query structure for reusable analytical pipelines.

(Click to view certificate)

• Outcome: reliable prompting patterns for LLM application building.
• Coverage:
  • Structured prompting for summarisation, inference, transformation, generation.
  • Iterative refinement patterns for reliability and consistency.
  • Practical OpenAI API exposure through course labs.

(Click to view certificate)

• Outcome: foundations of agentic workflows (plan/act/reflect) and tool-based systems.
• Coverage:
  • Task decomposition; reflection-driven improvement; evaluation loops.
  • Tool calling and API actions; retrieval/search patterns.
  • Multi-agent coordination patterns and practical lab-based implementations in Python.

(Click to view certificate)

• Outcome: graph-structured agent workflows with explicit control flow and tracing.
• Coverage:
  • Nodes/edges/state objects; deterministic orchestration and tool integration.
  • Retries and human-in-the-loop checkpoints; debugging via tracing.
  • Building complete agent pipelines in Python.

(Click to view certificate)

• Outcome: practical best practices for using Claude Code as a highly agentic coding assistant to plan, execute, and iterate on code with minimal supervision.
• Coverage:
  • Claude Code fundamentals: architecture, tool use, codebase navigation, and memory across sessions.
  • Context and memory discipline: supplying the right files or media, using escape, clear, and compact, and persisting project rules via CLAUDE.md.
  • Agentic workflow patterns: plan-first prompting, “thinking mode” for harder tasks, and subagents for ideation and parallel work.
  • Quality and iteration: test-writing, refactoring, and structured improvement loops to harden functionality.
  • Parallel delivery and integrations: multi-session development with git worktrees, plus GitHub issue and pull request workflows, hooks, and MCP servers such as Figma and Playwright applied to real builds.

(Click to view certificate)

• Outcome: hands-on ability to design, package, and deploy reusable “skills” (instruction folders) that make agents more reliable specialists across Anthropic’s ecosystem.
• Coverage:
  • Skills concept: turning general-purpose agents into specialists on demand; skill portability via an open standard.
  • Skill structure & format: skill folder anatomy, SKILL.md, and progressive disclosure for efficient context management.
  • Comparisons: when to use skills vs tools, MCP, and subagents.
  • Pre-built skills in practice: using Anthropic skills (Excel, PowerPoint, Skill Creation) in Claude.ai to build a marketing campaign analysis workflow.
  • Building custom skills (best practices): skills for generating practice questions from lecture notes; analyzing time-series data characteristics.
  • API integration: using custom + pre-built skills with the Claude API, including code execution and Files API to enable filesystem access and bash-driven Python execution.
  • Claude Code workflows: skill-driven code generation/review/testing pipelines; subagents with isolated context and specialized skills.
  • Agent SDK: building a research agent with the Claude Agent SDK that uses a skill to produce a learning guide from documentation, GitHub, and web search.

(Click to view certificate)

• Outcome: professional Git fluency for collaborative codebases.
• Coverage:
  • Core workflow: working tree → staging → commits; branching/merging; conflict resolution.
  • Inspecting change with git diff; reasoning across commits and branches.
  • Interrupt/recovery workflows: stash; restore, revert, reset; detached HEAD; reflogs.
  • Collaboration: remotes, push/pull/fetch, PRs, forks/clones; GitHub workflows.
  • Git internals mental model: objects (blobs/trees/commits) and annotated tags.
  • Markdown/READMEs, Gists, and GitHub Pages basics.
  • Rebasing and interactive rebase; tags and lightweight release patterns; custom Git aliases.

(Click to view certificate)

• Outcome: validated foundational AWS/cloud knowledge (conceptual understanding; not hands-on operations).
• Coverage (core services & concepts):
  • Compute & serverless: EC2, Lambda
  • Storage: S3, EBS, EFS
  • Databases: RDS, DynamoDB
  • Networking (high level): VPC, subnets, security groups, Internet Gateway, NAT Gateway
  • Security basics: IAM (roles/policies), shared responsibility model
  • Monitoring & audit: CloudWatch, CloudTrail
  • Cost & governance: pricing concepts, Cost Explorer, Budgets, AWS Organizations