AI
From models to intelligent systems.
A model is a component. Intelligence comes from how data, models, infrastructure and software work together.
The intelligence stack
Each layer depends on the one beneath it. Weakness at the bottom limits everything above.
- L1
Data
The raw material. Collected, cleaned, governed and made accessible.
- L2
Models
Statistical and neural models that learn patterns from data.
- L3
Infrastructure
Compute, serving and observability that run models reliably.
- L4
Applications
Software that puts model output in front of people and processes.
- L5
Agents
Systems that reason over goals and use tools to act.
- L6
Automation
End-to-end processes where intelligence runs continuously.
Areas
Where we focus
The disciplines that make up modern AI work.
- Generative AI
- Models that produce text, code, images and structured output — useful for drafting, summarising and transforming information.
- Machine learning
- Learning from historical data to classify, rank and estimate.
- AI agents
- Goal-directed systems that plan steps and call tools with defined permissions.
- NLP
- Understanding and generating language, including Arabic and bilingual content.
- Computer vision
- Interpreting images and video to detect, read and measure.
- Predictive AI
- Forecasting outcomes with explicit uncertainty.
- Model deployment
- Packaging, serving, versioning and monitoring models in production.