AI Academy

Disclaimer :
This page is still under construction.
It can have incomplete content.

Consult AI Experts in online-meeting on a weekly basis.

Dr. Thomas Meenken

>20 years of expertise in development of complex products

Rik Rasor

>10 years in Engineering IT
& 3 years in AI

Ruslan Bernijazov

>10 years in Software Engineering &
5 year in AI

Uwe Kloss

>15 years in Digital Transformation in
R&D Processes and Data

Subscribe to get all updates and information about the next meetings

Subscribe Now

Expert Call - Topics

Engage in our weekly expert sessions

Join our AI Academy and take part in regular online sessions with industry experts who answer your questions. Leverage the community to accelerate learning and drive innovation.
By joining the AI-Academy you get a continuous grow of AI competencies in your organization. Community learning will empower you to keep pace in the fast-changing world of AI. We will touch on the following topics:

A. Fundamentals & Current Models

Introduction to Transformer Architectures

Origin and core concept (Attention, Self-Attention)
Differences from earlier architectures (RNN, LSTM)
Relevance for modern LLMs

Deep Dive: Claude, Gemini, DeepSeek, Llama & Co.

Overview of key current Large Language Models
Technical differences and unique features
Application scenarios in industry

Memory in LLMs

How do LLMs store and process context over long conversations?
Technical concepts (Recurrent Attention, Memory Vectors, "Retrieval-Augmented Memory")
Use cases in engineering applications

Hallucinations in LLMs

What is a “hallucination” and how does it arise?
Impact on trustworthiness and data quality
Methods for measurement and mitigation

Bias in Data & Models

Typical forms of bias in training data
Effects on technical and ethical aspects
Strategies for preventing or correcting bias

B. Retrieval-Augmented Generation (RAG) & Knowledge Integration

RAG Fundamentals

Principle: Combining knowledge retrieval with generative AI
Typical architecture (Vector DB, Embeddings, Prompting)
Benefits and typical use cases in engineering

Knowledge Graphs & RAG

Using ontologies, taxonomies, and knowledge graphs
Ways to integrate domain-specific knowledge into LLMs
Examples from mechanical engineering, electronics, etc.

Multimodal RAG

RAG with images, CAD data, videos, or sensor data
Challenges in embedding generation for different modalities
Potential for quality control, documentation, technical diagnostics

Deep Research

Expanding RAG for deeper analysis (e.g., scientific papers, patents)
Automated literature analysis, state-of-the-art overview
Applications in R&D departments

C. Agentic AI & Automation

AI Agents in Engineering

Examples of automation in engineering processes (planning, simulation, design)
How can agents autonomously execute steps (e.g., invoking tools)?
Distinction from traditional workflow systems

Agentic AI Platforms (e.g., crewAI) & Copilot Studio

Practical examples of building and deploying agentic systems
Integration into existing engineering toolchains (e.g., CAD, PLM, ERP)
Advantages, limitations, and best practices

Guardrails in Engineering

Need for safety and quality control in AI agents
Governance approaches and technical implementations (e.g., prompt filters, policy checks)
Concrete examples of failure scenarios and how to prevent them

Manus System – Value of Tools for AI Agents

Introduction to "Manus System" (or similar tools) for agent control
Benefits for engineering teams: quality assurance, traceability, collaboration
Integration into MLOps/AIOps environments

D. Model Optimization & Deployment

Quantization & Compression

Overview of methods (Quantization, Pruning, Distillation)
Trade-offs between model size, accuracy, and performance
Practical examples of resource-efficient LLMs in engineering

Local LLM Deployment

Hardware and software requirements (GPU vs. CPU vs. TPU)
Privacy and compliance advantages of on-premises deployment
Common pitfalls and solutions (memory, scaling, maintenance)

LLM Fine-Tuning & Synthetic Data

Different fine-tuning approaches (Full, LoRA, Prompt Tuning)
Generating synthetic data for testing and training (e.g., simulation data)
Quality assurance and validation of synthetic datasets

Testing LLM Applications

Test strategies (unit tests for prompt templates, integration tests, red teaming)
Metrics for evaluating responses (Accuracy, F1, BLEU, etc.)
Automated test pipelines and Continuous Integration (CI)

E. Governance, Law & Security

EU AI Act & Consequences for Engineering

Overview of the planned regulation
Impact on development processes, documentation, compliance
Practical recommendations for preparation

Cybersecurity in the Context of LLMs

Attack vectors (Prompt Injection, Data Leakage, Model Inversion)
Protective measures and secure architectures
Relevance for mission-critical engineering data

Ethics & Responsibility (Responsible AI)

Why "Responsible AI" is particularly important in engineering
Risk assessment (e.g., in safety-critical systems, robotics)
Approaches for implementing ethical guidelines

F. Specific Application Fields in Engineering

Digital Twins & Simulation

AI-assisted simulations (e.g., Finite Element Methods, flow analysis)
Integration of real-time sensor data with digital twins
Optimization of development and operational processes

Generative Design & CAD

Automated design suggestions (e.g., topology optimization)
Use of LLMs for design documentation and variant management
Future outlook: AI as a co-designer

MLOps in Engineering

Building scalable and reproducible ML pipelines in industrial environments
Integration with PLM/ERP systems
Governance, monitoring, and continuous improvement

Edge AI & Embedded AI

Application scenarios in IoT and Industry 4.0
Hardware constraints and optimization strategies (quantization, model compression)
Live diagnostics and predictive maintenance

New Material Development & Additive Manufacturing