Four layers of the AI ecosystem
Infrastructure
Chips, data centres, cloud, data and development platforms.
Foundation models
Providers and research teams for language, image or multimodal models.
Applications
Software for learning, medicine, industry, media, administration or coding.
Integration
Consulting, security, data preparation and workflow implementation.
Many well-known companies operate across several layers. This is a thinking model, not a fixed industry list.
Analyse companies meaningfully
- Which concrete problem does the product solve?
- Which data, models and infrastructure are required?
- How is quality measured and which errors remain?
- Who pays, who uses and who carries responsibility?
- Which human tasks emerge, and which change?
For students, this task perspective is often more useful than a list of current market values.
Turn company research into a learning project
- compare two products by benefit, data and risk
- draw a value chain
- investigate one career role in an AI company
- test a product claim through evidence or experiments
Current models, tools and company areas are available in the KAI hub.