Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Course Outline
Introduction to GPT-5 and Developer Capabilities
- Key capabilities of GPT-5, including multi-modality and agent features.
- Guidance on choosing models, understanding pricing, and managing limits.
- Ethical considerations and enterprise governance.
Prompting and System Design for Reliable Outputs
- Prompt patterns, system messages, and context engineering.
- Comparison of chain-of-thought versus concise prompting, along with few-shot techniques.
- Testing prompts and establishing clear acceptance criteria.
APIs, SDKs, and Local Dev Workflow
- Calling GPT-5 APIs, utilising SDKs, and managing authentication and secrets.
- Local development practices, mocking responses, and sandboxing.
- Versioning, request/response schemas, and error handling.
Building Agents and Tool Integrations
- Designing safe agent architectures and tool interfaces.
- Strategies for routing, orchestration, and fallback mechanisms.
- Managing rate-limits, concurrency control, and transactional considerations.
Testing, Evaluation and Validation
- Developing automated test suites for prompts and behaviours.
- Conducting red-teaming, fuzz testing, and analysing adversarial examples.
- Measuring accuracy, hallucination rates, and user satisfaction.
Deployment, Monitoring and Observability
- Implementing CI/CD patterns for model-enabled features and canary releases.
- Logging, tracing, and telemetry for prompt-level observability.
- Alerting, SLA considerations, and incident response protocols.
Security, Privacy and Cost Optimization
- Data handling, PI/PHI considerations, and context sanitisation.
- Access control, auditing, and compliance checkpoints.
- Optimising token usage, batching, and caching strategies.
Summary and Next Steps
Requirements
- A solid understanding of at least one programming language, such as Python or JavaScript.
- Experience in calling REST APIs or using SDKs.
- Basic familiarity with ML/AI concepts and JSON data structures.
Audience
- Software engineers
- ML engineers
- DevOps / SRE engineers
14 Hours
Testimonials (1)
Able to pivot upon audience suggestions - ie able to create a real AI agent scenario on the spot.
