AI Skills Guides for UK Engineers
2026 Edition
Twenty deep technical guides covering the skills that matter most for AI and ML engineering careers in the UK — from Python fundamentals to fine-tuning LLMs, MLOps infrastructure, and responsible AI. Each guide is written by engineers, not marketers.
Core ML
3 guidesPython for Machine Learning
NumPy, Pandas, scikit-learn, type hints, testing, and production Python patterns for ML engineers.
PyTorch for Deep Learning
Tensors, autograd, nn.Module, training loops, mixed precision, DDP, and model serialisation.
Transformer Architecture
Self-attention, multi-head attention, RoPE, GQA, scaling laws, and model family comparisons.
GenAI
6 guidesFine-tuning Large Language Models
LoRA, QLoRA, SFT, DPO — every PEFT and alignment technique with practical implementation guidance.
Retrieval-Augmented Generation (RAG)
Chunking, embeddings, vector search, reranking, HyDE, RAGAS evaluation, and production RAG patterns.
LangChain and AI Agents
LCEL, ReAct agents, tool use, LangGraph for stateful workflows, and production agent architecture.
Vector Databases
HNSW indexing, similarity metrics, Pinecone vs Qdrant vs Weaviate vs pgvector, hybrid search.
HuggingFace Transformers
Transformers, Datasets, PEFT, Accelerate, TRL, and the Hub ecosystem for applied LLM work.
RLHF and LLM Alignment
Reward modelling, PPO, GRPO, DPO, Constitutional AI, preference data collection, and TRL.
MLOps
6 guidesMLflow Experiment Tracking
Tracking, Projects, Models, and the Model Registry — the open-source MLOps platform UK teams rely on.
Docker for Machine Learning
Efficient ML Dockerfiles, GPU containers with CUDA, multi-stage builds, and FastAPI model serving.
AWS SageMaker
Training Jobs, Endpoints, Batch Transform, Pipelines, Feature Store, and cost optimisation.
FastAPI for Model Deployment
Async endpoints, Pydantic validation, model loading at startup, LLM streaming, API versioning.
Weights & Biases (W&B)
Experiment tracking, Sweeps for hyperparameter optimisation, Artifacts, and Reports.
Kubernetes for MLOps
GPU scheduling, Kubeflow, KServe for model serving, autoscaling, and resource management.
Data Engineering
2 guidesSQL for Data Science
Window functions, CTEs, feature engineering, point-in-time correctness, BigQuery, Snowflake, dbt.
Apache Spark for Machine Learning
PySpark DataFrames, MLlib Pipeline API, Delta Lake for feature engineering, Databricks platform.
Specialisms
3 guidesComputer Vision Skills
CNN architectures, YOLO object detection, Vision Transformers, OpenCV, Albumentations, mAP.
NLP Engineering Skills
Tokenisation, NER, semantic similarity, spaCy, HuggingFace, BLEU, ROUGE, and BERTScore.
Responsible AI and AI Ethics
UK AI regulation, fairness metrics, SHAP explainability, model cards, and differential privacy.