Track
AI Architect
“I want to design AI systems at organisation scale.”
The longest line. Assumes you will build first, then take responsibility for boundaries, cost, governance and failure modes.
- Python
The only genuinely non-negotiable prerequisite in the entire Atlas.
- Math for AI
The specific mathematics that appears in practice, and nothing beyond it.
- Git & Collaboration
Version control, review and reproducibility. Assumed silently by every team.
- Linear Algebra
Vectors, matrices and the operations every model is secretly made of.
- Calculus & Optimisation
Derivatives, the chain rule, and what it means to walk downhill in a loss landscape.
- Machine Learning
Systems whose behaviour is learned from data rather than written as rules.
- Gradient Descent
The optimisation procedure underneath essentially every model in this Atlas.
- Deep Learning
Many-layered neural networks that learn their own features instead of being handed them.
- Neural Networks
Layers, weights and activations. The unit of construction for everything downstream.
- Attention
Letting every position look at every other position. The hinge the modern field turns on.
- Transformers
Attention, feed-forward layers, residuals and normalisation. The block that ate the field.
- Large Language Models
Transformers trained on enough text to become general-purpose reasoning surfaces.
- Embeddings
Meaning as coordinates. The representation that makes search, clustering and RAG possible.
- Prompt Engineering
Specifying a task precisely enough that a probabilistic system does it reliably.
- RAG
Give the model the right documents at query time instead of hoping it memorised them.
- AI Agents
Models that decide what to do next, call tools, and act in a loop rather than answering once.
- AI Engineering
The bridge from a working notebook to a system real users depend on.
- Probability & Statistics
Distributions, expectation, and why every evaluation number needs an error bar.
- Supervised Learning
Learning a mapping from labelled examples. The workhorse of applied ML.
- Model Evaluation
The skill that separates people who ship models from people who publish notebooks.
- LLM Evaluation
Measuring quality when there is no single correct output. The hardest unsolved problem in shipping.
- Observability & Evals
Tracing, logging and regression suites. Knowing you broke it before a user tells you.
- Model Serving
APIs, batching, streaming, and the deployment surface between model and product.
- Inference Optimisation
Quantisation, caching, speculative decoding. Buying back latency and margin.
- Cost & Latency
Token economics, model routing and caching. The constraint that decides architecture.
- SQL
Where the training data actually lives, and the fastest way to interrogate it.
- Data Wrangling
Loading, cleaning and reshaping data. Realistically most of the job.
- Guardrails & Safety
Input validation, output filtering, prompt injection, and where to put the boundary.
- Data Strategy
Provenance, licensing, freshness and permissions. The unglamorous moat.
- AI Architecture
System-level design: model routing, boundaries, failure modes, governance. The summit node.