AI Training Catalog

Understand how AI and machine-learning systems are built, where they fail, and the new attack surface they introduce to an organisation.

Equip non-technical staff to use AI safely and productively — capabilities, limitations, data handling and acceptable use.

Build AI that is fair, transparent and accountable — addressing bias, explainability and human oversight across the lifecycle.

Understand how modern generative AI works and how to apply it responsibly and effectively across real business problems.

Master the craft of prompting — structured prompts, context, evaluation and patterns that get reliable results from LLMs.

Build a working foundation in machine learning — core algorithms, the modelling workflow and how to evaluate results honestly.

Establish governance for AI systems — risk classification, assurance, documentation and alignment to emerging AI regulation and standards.

Operationalise the NIST AI Risk Management Framework — Govern, Map, Measure and Manage — across your AI portfolio.

Navigate the EU AI Act and the emerging global patchwork of AI regulation, and translate obligations into practical compliance.

Defend production GenAI systems against prompt injection and jailbreaks with layered guardrails, input/output controls and monitoring.

Secure the machine-learning supply chain and operations — data pipelines, model registries, deployment and monitoring.

Protect the data that powers AI — governance, privacy-preserving learning and controls across collection, training and inference.

Apply machine learning to strengthen defence — anomaly detection, alert triage and automation — while understanding its limits and risks.

Secure retrieval-augmented generation systems — protecting vector stores, retrieval pipelines and the data they expose to models.

Govern models across their lifecycle — model cards, registries, approval gates, monitoring and decommissioning — for accountable AI.

Build production-grade applications on top of large language models — orchestration, tools, evaluation and deployment.

Design and build RAG systems that ground LLMs in your own data — chunking, embeddings, retrieval and evaluation.

Work with embeddings and vector databases — the backbone of semantic search, RAG and recommendation at scale.

Understand and build neural networks — architectures, training dynamics and the practical craft of deep learning.

Process and understand human language with modern NLP — from embeddings and transformers to real text applications.

Build systems that see — image classification, detection and segmentation with modern deep-learning vision models.

Build the data foundation AI depends on — pipelines, quality, feature stores and governance for trustworthy models.

Make AI decisions understandable and defensible — interpretability methods, model transparency and explanations for stakeholders.

Design and implement an AI management system to ISO/IEC 42001 — the international standard for responsible AI governance.

Shape responsible national AI policy — balancing innovation, rights and security with practical governance for the public sector.

Attack and defend machine-learning models against evasion, poisoning, extraction and inference attacks using hands-on adversarial techniques.

Secure LLM and generative-AI applications against prompt injection, data leakage and abuse, and design safe retrieval and agent architectures.

Systematically attack large language models and GenAI applications to surface safety, security and misuse failures before adversaries do.

Probe AI systems for safety, security and misuse failures, and build repeatable model-assurance evaluation pipelines.

Detect, investigate and recover from incidents targeting AI systems — model compromise, data poisoning and pipeline tampering.

Secure the AI supply chain — third-party models, datasets and dependencies — against poisoning, backdoors and provenance gaps.

Establish trust in AI outputs and models through watermarking, content provenance (C2PA) and verifiable model lineage.

Detect and counter AI-generated deepfakes and influence operations that threaten institutions, elections and public trust.

Build and operate machine-learning detection at scale — feature engineering, model selection and evaluation for real security telemetry.

Secure autonomous and tool-using AI agents — controlling permissions, tool access, memory and the new risks of agentic workflows.

Design AI systems that are secure and trustworthy by construction — applying Google SAIF and security patterns across the stack.

Adapt foundation models to your domain — fine-tuning, parameter-efficient methods (LoRA) and alignment for specialised tasks.

Operate LLMs in production — serving, scaling, observability, evaluation and cost control across the model lifecycle.

Build systems that reason across text, image, audio and video using modern multimodal models and pipelines.

Design autonomous and multi-agent systems — planning, tool use, memory and orchestration — for complex real-world tasks.

Rigorously evaluate LLMs and GenAI systems — building evals, benchmarks and guardrails that measure quality, safety and regression.

Operationalise machine learning at scale — pipelines, CI/CD for models, monitoring and reproducibility across the lifecycle.

Squeeze real performance from models — feature engineering, hyperparameter tuning, compression and efficient inference.

Train agents that learn by interaction — the foundations of reinforcement learning and its modern applications, including RLHF.

Train and deploy models without exposing sensitive data — federated learning, differential privacy and secure computation.

Detect, measure and mitigate bias in AI systems, and conduct independent algorithmic audits for fairness and accountability.

Track and operationalise intelligence on attacks against AI systems using the MITRE ATLAS knowledge base of real-world adversary techniques.

Adopt AI safely in critical national systems — balancing capability with the assurance, resilience and oversight these systems demand.

A board-level briefing on capturing AI's value while governing its strategic, security and societal risks responsibly.

Build national AI capability and sovereignty — strategy, infrastructure, talent and governance to lead responsibly at scale.

A top-tier mastery programme for leaders to shape enterprise AI strategy, nation-scale governance frameworks, and responsible AI transformation leadership.

Expert-level mastery to design, secure, and operate production-scale AI systems end-to-end, including architecture, advanced LLMOps, and model assurance.

A mastery programme for power users to lead cross-unit AI adoption, design AI solutions, and cultivate a responsible innovation culture.