Practical notes
Articles · ai
Articles on data platforms, applications, AI and cloud, explaining technical mechanisms and architecture decisions.
A data platform for analytics, applications and AI
Data models, history and publishing interfaces that connect a data warehouse to analytical queries, applications and predictive models.
Four questions about the EU AI Act for a Swiss company
The company’s role, its EU connection and the system’s purpose determine the assessment. Those conclusions must lead to design decisions the team can verify.
Designing private AI for law firms around data access and verification
Confidentiality, permissions and verification guide the architecture. A server is one implementation option within that assessment.
A versioned knowledge base for teams and AI assistants
Approved sources, their history and controlled updates determine a knowledge base’s reliability. Text files can make that chain easier to inspect and maintain.
Choosing agent memory and managing its lifecycle
Conversation history, reference documents and persistent memory serve different purposes. Their value depends on provenance, permissions and the ability to correct information.
Reducing IT costs through a complete cost comparison
The dominant expense, the effort of change and the operating cost determine the saving. A comparison method, with AI as a worked example.
Designing retrieval, tools and controls for an AI agent
RAG, loops and graphs address different decisions. One supplier dossier shows their roles and the conditions that justify each component.
Evaluating open-weight model quality, availability and operating cost
Public rankings identify candidates. Selection depends on the licence, performance on the task and the resources required to operate the model.
Four AI agent workflows and their acceptance criteria
An agent connects information analysis with an organisation’s tools. Four examples show the integrations, expected results and checks needed to evaluate that work.
Organising AI tools, data and responsibilities at school
An educational use becomes practical when the school defines its purpose, permitted data, accounts and review rules.
Three configurations for sizing local AI
Memory, throughput and concurrent demand guide the hardware choice. Three profiles help prepare a trial and compare total operating cost.
Deploying internal AI with open models and suitable infrastructure
Required quality, data constraints and operating capacity determine the deployment model. The model and its environment need to be evaluated together.
Choosing an AI architecture for health data and research results
Patient records and unpublished findings can coexist in one organisation. Their uses and disclosure rules determine the controls the system needs.