Building with AI
MLOps and LLMOps course
AI in production, with the same rigour as the rest of your software.
Training a model or building an LLM demo is only the beginning. The real work starts when the system is in production: versions, deployments, monitoring, drift and costs. MLOps and LLMOps are the practices that make AI work as well on day one hundred as on day one.
This course covers both worlds: the classic machine learning cycle (experiments, model registry, pipelines, serving and drift) and what is specific to LLM applications (versioned prompts, tracing, continuous evaluation and cost control). With a pipeline set up from start to finish.
What you will learn
- Version data, models, prompts and experiments
- Automate training, evaluation and deployment
- Serve models at scale
- Monitor quality, drift, cost and latency
- Manage the lifecycle of LLM applications
Syllabus
8 modules, from the fundamentals to the final project.
-
The AI lifecycle
- From notebook to service
- MLOps versus LLMOps
- Maturity levels
-
Experiments and versions
- Experiment tracking and model registry
- Data versioning
- Reproducibility
-
Pipelines
- Automated training and evaluation
- Continuous integration and deployment for models
-
Serving models
- Inference APIs and containers
- Scaling, GPUs and batch processing
- Inference servers for LLMs
-
Monitoring
- Technical and business metrics
- Data and prediction drift
- Alerts and retraining
-
LLMOps
- Prompt management and versioning
- Tracing and observability for LLM applications
- Continuous evaluation
-
Costs
- Cost per request and per user
- Caching, smaller models and routing
-
Governance
- Traceability for audits
- Model documentation and AI Act requirements
Tools you will practise with
- MLflow
- DVC
- Docker
- GitHub Actions
- vLLM
- Langfuse
Hands-on project
You set up the pipeline for a sample model, from training and registration to containerised deployment with monitoring, and add the observability and evaluation layer for an LLM application.
Requirements
- Python programming.
- Knowledge of Docker and Git.
- Machine learning experience recommended.
Other courses
-
AI in apps and digital products course
AI features users understand, use and trust.
- Duration: 12 hours
- Level: Intermediate
-
AI-assisted programming course
Code faster with coding assistants and agents, without losing quality.
- Duration: 12 hours
- Level: Intermediate
-
Generative AI system evaluation course
Measure your AI application's quality with data, not impressions.
- Duration: 12 hours
- Level: Advanced
Where it is applied
-
Custom AI development
We design and build artificial intelligence solutions tailored to your processes, data and tools when an off-the-shelf solution does not fit or falls short.
Want this course for your team?
We adapt the content, length and format to your company.