AI training for your team

For whom:

This module is designed for companies that want to strategically integrate AI into their business processes but have not yet delved deeply into AI applications and frameworks.

Content:

• Fundamentals, common hurdles, and how to overcome them – what should be avoided?
• What constitutes a holistic AI strategy?
• Compliance: What needs to be considered when working with LLMs, especially regarding data transmission?
• Prompt engineering: Basics, advanced techniques, and collaborative solutions
• Integrating AI into products: Value assessment and implementation strategies
• Incorporating your own data into AI using Retrieval-Augmented Generation (RAG)

Outcomes:

• A tailored AI strategy for the company
• Overview of compliance and security considerations
• Best practices for integrating AI technologies into business operations

For whom:

This module is aimed at companies that are already familiar with the basics of AI and are now looking to implement AI in specific processes or products.

Content:

• Analysis of the current business model, existing processes, and available data
• Identification of potential AI use cases based on business goals and industry-specific challenges
• Evaluation of use cases in terms of value, feasibility, data protection/compliance, and cost-effectiveness
• Introduction to suitable implementation technologies, including a preliminary cost and time estimate
• Derivation of a strategic roadmap for AI implementation

Outcomes:

• Documentation of identified and prioritized AI use cases
• Cost and resource assessment
• Strategic road map with recommendations for proof of concepts or pilot projects for AI integration

For whom:

This module is designed for companies that have already identified concrete AI use cases and now want to build their own locally operated Large Language Model (LLM) – due to data sovereignty concerns or specific requirements for security and customization.

Content:

Technical foundations of LLMs

• Semantic vector space / embeddings
• Attention
• Context window
• System prompts
• Reasoning models
• Pretraining + fine-tuning

Local LLMs with Ollama

• Which models for which use cases
• Integration into existing systems: APIs, access concepts, user roles
• Infrastructure: What fits your needs? (On-premises, private cloud, hybrid models)
• Security and data protection requirements for local LLMs
• Proof of concept and example implementation

Incorporating your own data into the AI using RAG: Retrieval-Augmented Generation

• How it works
• Embeddings and vector databases
• Interaction between prompt and RAG
• Use cases and examples

Best practices and “what doesn’t work”

• ChatGPT
• aider, zed, Cursor
• Testing & refactoring
• Use in client projects (local and OpenAI-based)

Outcomes:

• Architecture concept for a local LLM based on your company’s requirements
• Recommendations for suitable models, tools, and frameworks
• Concrete implementation steps for building a production-ready LLM stack
• Overview of security and compliance requirements when operating your own models