Learning Objectives
By the end of this training programme, you will be able to:
Technical Competencies
Prompt Engineering
- Master the AUTOMAT Framework for structured scientific prompts
- Apply the CO-STAR Framework for context-rich communication
- Design high-fidelity outputs suitable for rigorous scientific enquiry
- Engineer constraints to focus AI on specific scientific domains
Model Understanding
- Distinguish between NLMs and LLMs and select appropriate tools
- Understand hallucination mechanisms and implement verification protocols
- Navigate context windows for complex scientific tasks
- Optimise token usage for efficient inference
Operational Competencies
Security & Compliance
- Identify sensitive data that must never be shared with public models
- Apply the Red List protocol for IP protection
- Sanitise data before processing with external tools
- Implement air-gapped workflows for confidential research
Sustainability
- Measure carbon footprint of AI workflows
- Apply Green AI principles to minimise energy consumption
- Batch queries efficiently to reduce computational waste
- Choose appropriate model sizes for different tasks
Practical Applications
Materials Science Workflows
- Automate literature synthesis for polymer chemistry and nanotechnology
- Format synthesis protocols in standardised templates
- Generate data analysis scripts for tensile testing and SEM imaging
- Extract structured data from unstructured research documents
Quality Assurance
- Verify AI outputs against authoritative sources
- Detect and correct hallucinations in technical content
- Audit AI-generated reports for scientific accuracy
- Implement human-in-the-loop workflows
Ethical Awareness
- Recognise bias in AI outputs
- Understand data rights and training data provenance
- Navigate environmental implications of AI deployment
- Apply responsible AI principles in daily workflows
Assessment Criteria
Throughout the workshop, you'll demonstrate these competencies through:
- Interactive Exercises: "The Hallucination Hunt", "Green Optimisation Challenge"
- Real-World Applications: Processing actual tensile test data and SEM images
- Prompt Refinement: Iterative improvement of scientific queries
- Peer Review: Evaluating and improving colleague's prompts
Next: About This Course →