The real tools behind our courses

No toy examples on made-up software. Here is the actual infrastructure and third-party components each course is built on — read live from our infrastructure catalogue, not a marketing list.

Containers & OrchestrationDevOps

DockerOptional

Guides: A checklist for training a neural network that works, A minimal evaluation harness for LLM apps, Document AI: OCR, extraction and validation, MLOps for small teams: what actually matters

KubernetesOptional

Guides: Monitoring a deployed model: four layers, The KV cache: why LLM serving is a memory problem

Monitoring & ObservabilityDevOps

PrometheusRecommended

Guides: MLOps for small teams: what actually matters, Monitoring a deployed model: four layers, Prompt injection: what it is and how to reduce the risk

GrafanaOptional

Guides: MLOps for small teams: what actually matters, Monitoring a deployed model: four layers

ElasticsearchOptional

Guides: Chunking documents so retrieval works

OpenTelemetryOptional

Guides: How to tell whether your RAG assistant is any good, Monitoring a deployed model: four layers

Messaging & Event StreamingDevOps

Apache KafkaOptional

Guides: Idempotent pipelines: safe to re-run

Backend FrameworksPaaS

FastAPIOptional

Guides: Document AI: OCR, extraction and validation, Never trust the model's JSON: validate and repair

Databases — Relational (SQL)Data

PostgreSQLOptional

Guides: Document AI: OCR, extraction and validation, Idempotent pipelines: safe to re-run, What is an embedding? Meaning as numbers

Vector DatabasesAIaaS

ChromaDBOptional

Used in Introduction to Retrieval-Augmented Generation (RAG)
Guides: What is an embedding? Meaning as numbers

PineconeRequired

Used in Introduction to Retrieval-Augmented Generation (RAG)

pgvectorOptional

Guides: Chunking documents so retrieval works, How to choose between RAG and fine-tuning, RAG, fine-tuning or prompting? A decision guide, What is an embedding? Meaning as numbers

FAISSOptional

Guides: Chunking documents so retrieval works, How to choose between RAG and fine-tuning, What is an embedding? Meaning as numbers

QdrantOptional

Guides: What is an embedding? Meaning as numbers

MilvusOptional

Guides: What is an embedding? Meaning as numbers

Object & File StorageIaaS

MinIOOptional

Guides: Document AI: OCR, extraction and validation

Data EngineeringMLOps

Apache AirflowOptional

Guides: A minimal evaluation harness for LLM apps, Idempotent pipelines: safe to re-run, MLOps for small teams: what actually matters

Apache SparkOptional

Guides: Idempotent pipelines: safe to re-run

dbtRecommended

Guides: Idempotent pipelines: safe to re-run

PrefectOptional

Guides: Idempotent pipelines: safe to re-run

DagsterOptional

Guides: Idempotent pipelines: safe to re-run

Delta LakeRecommended

Guides: Idempotent pipelines: safe to re-run

Apache IcebergOptional

Guides: Idempotent pipelines: safe to re-run

ML FrameworksAIaaS

Hugging FaceOptional

Guides: Document AI: OCR, extraction and validation, How text becomes tokens (BPE in plain words), How to choose between RAG and fine-tuning, Python and maths for ML: where to start

Scikit-learnRequired

Guides: Data leakage: the reason a great model fails in production, Precision, recall and choosing the right metric, Python and maths for ML: where to start, Train, validation and test sets: why three?

PyTorchRequired

Guides: A checklist for training a neural network that works, Document AI: OCR, extraction and validation, How text becomes tokens (BPE in plain words), IoU, precision–recall and mAP for object detection

OpenCVRecommended

Guides: Document AI: OCR, extraction and validation, IoU, precision–recall and mAP for object detection

ONNXOptional

Guides: A checklist for training a neural network that works, IoU, precision–recall and mAP for object detection

TensorFlowOptional

Guides: A checklist for training a neural network that works

MLOps PlatformsMLOps

MLflowOptional

Guides: A checklist for training a neural network that works, A minimal evaluation harness for LLM apps, Data leakage: the reason a great model fails in production, How to tell whether your RAG assistant is any good

Weights & BiasesOptional

Guides: A checklist for training a neural network that works, A minimal evaluation harness for LLM apps, When fine-tuning (and LoRA) is worth it

DVCRecommended

Guides: Data leakage: the reason a great model fails in production, MLOps for small teams: what actually matters, Train, validation and test sets: why three?

Seldon CoreOptional

Guides: Monitoring a deployed model: four layers

RayOptional

Guides: The KV cache: why LLM serving is a memory problem

AI / LLM ProvidersAIaaS

AnthropicOptional

Used in Introduction to Retrieval-Augmented Generation (RAG)

OllamaOptional

Used in Introduction to Retrieval-Augmented Generation (RAG)
Guides: The KV cache: why LLM serving is a memory problem

OpenAIRequired

Used in Introduction to Retrieval-Augmented Generation (RAG)
Guides: How text becomes tokens (BPE in plain words), How to choose between RAG and fine-tuning, Never trust the model's JSON: validate and repair, Prompt injection: what it is and how to reduce the risk

LLM & GenAI FrameworksAIaaS

LangChainRecommended

Used in Introduction to Retrieval-Augmented Generation (RAG)
Guides: Chunking documents so retrieval works, How to tell whether your RAG assistant is any good, Never trust the model's JSON: validate and repair, Prompt injection: what it is and how to reduce the risk

LlamaIndexRecommended

Guides: Chunking documents so retrieval works, How to choose between RAG and fine-tuning, How to tell whether your RAG assistant is any good, RAG, fine-tuning or prompting? A decision guide

vLLMOptional

Guides: The KV cache: why LLM serving is a memory problem, When fine-tuning (and LoRA) is worth it