LLM Engineering & Assurance
A hands-on course in building, deploying, integrating and assuring large language models.
Invitation only
Peek at free lessons now — full access arrives by invitation.
Explore what’s unlocked below. When you’re invited, the rest of the course opens up.
LLM Engineering: From Scratch to Assurance is a hands-on course designed to give learners a practical understanding of how Large Language Models work, how they are built, and how they can be deployed in real-world applications. Starting from the fundamentals of tokens, embeddings, attention, and transformers, participants will build a small language model from scratch and gain an intuitive understanding of the key components behind modern LLMs.
The course then moves from theory into engineering practice, covering model inference, deployment, APIs, performance optimisation, structured outputs, and tool calling. Through practical exercises, learners will build and integrate LLM-powered systems, progressing from a minimal model to a functional application that can interact with external tools and services.
The final part focuses on evaluating and assuring LLM systems for reliable and responsible deployment. Learners will explore hallucination, robustness, security, adversarial testing, red teaming, and evaluation techniques, with applications across defence, government, healthcare, finance, and other regulated or safety-critical sectors. The course concludes with a hands-on capstone where participants build, deploy, evaluate, and assure a complete LLM system.
No learning outcomes listed yet.
- Basic programming experience
- preferably with Python.
- Basic understanding of machine learning concepts.
- Familiarity with Python libraries such as NumPy and PyTorch is helpful but not required.
- Basic knowledge of neural networks is recommended.
- Basic command-line and Git experience is helpful.
- No prior experience building or training LLMs is required.