Skip to content

About This Course

Course Designer

These course materials have been designed and developed by Avgi Stavrou, as part of AmaDema's commitment to upskilling its R&D workforce in AI technologies whilst maintaining the highest standards of intellectual property protection and environmental responsibility.

Training Philosophy

This programme differs from generic AI training in three fundamental ways:

1. Materials Science Focus

Every example, exercise, and framework application is tailored to the specific needs of materials scientists and nanotechnology researchers. You won't find generic "write an email" prompts here, instead, you'll learn to:

  • Extract polymer synthesis parameters from literature
  • Analyse tensile testing data for non-oxide ceramics
  • Generate structured reports for IP legal teams
  • Automate SEM image analysis workflows

2. Security-First Approach

We operate under the principle that access to AI tools does not require data exposure. You'll learn to:

  • Identify the "Red List" of data that must never be uploaded
  • Sanitise datasets before external processing
  • Use local, air-gapped models for sensitive work
  • Balance utility with security

3. Environmental Consciousness

AI is not free; computationally or environmentally. This course treats Green AI as a core operational requirement, not an optional consideration. You'll understand the carbon cost of your queries and learn to optimise for both quality and sustainability.

Teaching Materials & Deliverables

This course includes comprehensive support materials for long-term adoption:

📘 The Prompt Engineering Cheat Sheet

A "book of spells" for the R&D team, containing:

  • Visual templates for AUTOMAT and CO-STAR frameworks
  • Pre-tested "Golden Prompts" for common lab tasks
  • The "Red List" for data security
  • Quick reference for model selection

🔬 Interactive Nano-Sandbox

A secure, local instance of Llama 3.2 (3 billion parameters) that allows you to:

  • Practice with real, sensitive data during the workshop
  • Experiment in a completely offline environment
  • Eliminate IP risk whilst learning
  • Test prompts without external data exposure

💻 Practical Exercises

Highly interactive sessions including:

  • The Hallucination Hunt: Audit an AI-generated chemical report to find deliberate errors
  • Green Optimisation Challenge: Refactor inefficient prompt chains into lean instructions
  • Real Data Processing: Work with actual tensile test data and SEM images
  • Prompt Peer Review: Evaluate and improve colleague's prompts

📦 Take-Home Resources

  • Downloadable PDFs of frameworks and cheat sheets
  • Template prompts for common R&D tasks

Course Structure

Format: 4-day intensive workshop (1.5 hours per day)
Target Audience: Material Scientists and R&D staff
Prerequisites: None, presumes no prior AI experience
Environment: Hands-on with local sandbox (no installation required)

Session Breakdown

Day Focus Duration Activities
1-2 Precision Engineering & Security 3 hours AUTOMAT/CO-STAR frameworks, Red List protocol, Hallucination Hunt
3-4 Technical Architecture & Sustainability 3 hours NLM vs LLM, Green AI, Optimisation Challenge, Real data processing

Support & Feedback

For queries related to this course, technical issues, or suggestions for improvement:

Contact: Avgi Stavrou
Email: Avgi Stavrou AMA Ask Me Anything

Ready to start? Head to What are LLMs? →