How a large language model (LLM) works
- Architecture of modern LLMs in simple terms: tokens, context window, and the “needle in a haystack” problem
- API basics: what an Application Programming Interface is and how to use it
- Technical details of LLMs: temperature, Top-P (nucleus sampling), practical examples, and working with hallucinations
- APIs, pricing, and ready-made services: how to launch an AI project quickly
- AI agents, Command-Line Interface (CLI), and services for automating business processes
- Context enrichment with MCP
Comparison of large language models: Claude, Gemini, and others
- GPT-OSS, LLaMA, and other models: when to choose open-source projects and when to choose commercial ones
- How to download and use free language models
- The US and China: distinctive features of language models
Hands-on practice in other useful AI services
- Creating websites with AI without programming
- Searching for information with AI
- Learning with AI
- Creating presentations and charts with AI
- Generating video and audio with AI
Methods for optimising AI for business needs (without going deep into code)
- Distillation, quantisation
- Important abbreviations for language models
- Retrieval-augmented generation (RAG)
- Optimisation for company-specific tasks: fine-tuning
- Vectorisation and useful services for business
- Processing speed and specialised chips for AI
- Tests for language models
- Agenda for C-level executives
- Risks and barriers
- AI implementation opportunities: MVP projects