Training Catalogue / Security and Threat Modelling
Architecting Secure LLM Applications
Modules
Take single modules or the whole track. Two hour sessions, combined into half days where it suits your calendar.
Secure-by-Design for AI Features
Build security into AI features from the start, not bolted on afterwards.
- The LLM security landscape and OWASP LLM Top 10
- Trust boundaries, zero trust and least privilege for AI
- The unique attack vectors LLM integration introduces
- Where AI Threat Modelling finds risks, this fixes them
Defending the AI Data and Tool Flow
Prompt-injection defence, safe data handling and secure tool calling.
- Prompt-injection defence and input sanitisation
- Secure data handling and training-data leakage prevention
- Output handling, validation and sandboxing
- RAG and tool-calling security
Hardening AI for Production
Rate limiting, monitoring and the supply-chain concerns of shipping AI.
- Supply-chain and deployment security
- API gateways, rate limiting and circuit breakers
- Logging, monitoring and detection
- Security testing, red teaming and production hardening
Teams Often Pair This With
Related modules other teams add alongside this track.
Securing RAG and Agentic AI
Retrieval poisoning, autonomous-action risk and multi-agent coordination threats.
View module Applied AI Engineering · Building GenAI ApplicationsShipping AI Features to Production
Scaling, caching, monitoring and the operational realities of running AI features.
View moduleEnquire About This Training
Tell us which modules interest you and we will propose a plan within 24 hours. Choosing Enquire on a module above fills in the subject for you. Add more than one if you like.