Glossary
Glossary for AI and digital product development
Short, precise explanations of the terms we use most often — from AI integration and RAG to EHF and GEO. Adapted to Norwegian business context.
- AI agent
- An AI agent is software that sets goals, plans steps, and acts autonomously to achieve them — without continuous human oversight.
- AI integration
- AI integration is the process of wiring language models, RAG, or predictive models directly into a business's existing systems and workflows.
- API integration
- An API integration connects two systems through their programming interfaces (APIs), so data flows between them automatically instead of being typed in twice.
- As-built documentation
- As-built documentation shows the equipment or installation as it was actually built and installed, with every change from the original drawings included.
- Billable utilisation
- Billable utilisation is the share of working hours invoiced to customers, calculated as billable hours divided by the total hours logged.
- EHF (Elektronisk HandelsFormat)
- EHF is the Norwegian standard for electronic invoices — mandatory for any supplier to the Norwegian public sector.
- Embedding
- An embedding is a numerical representation of text (or image) that lets AI systems compare semantic similarity.
- Engineer-to-order
- Engineer-to-order (ETO) is a production model in which the product is designed for each order to the customer's specification, before materials are bought and the equipment is built.
- FAT/SAT
- A FAT (Factory Acceptance Test) is the test in which the customer accepts the equipment at the supplier's premises before shipping. A SAT (Site Acceptance Test) is the test after installation, where the equipment is tried in its real environment.
- Fine-tuning
- Fine-tuning is the process of further training an existing AI model on specific data to improve performance for a narrow use case.
- Forward deployed engineer
- A forward deployed engineer (FDE) is a developer who works at the customer's site, not just from their own office, finds the real problem together with the people who do the work, and builds the solution there.
- GEO / AEO
- GEO and AEO are disciplines for getting cited or referenced by AI search engines like ChatGPT, Perplexity, and Google AI Overviews.
- LLM (Large Language Model)
- An LLM is a large language model trained on enormous text volumes that can generate, summarise, and analyse text in a human-like way.
- Manufacturing Record Book
- A Manufacturing Record Book (MRB) is the complete manufacturing documentation for a delivery: certificates, welding and NDT records, test and dimensional reports that show the equipment was made to requirements.
- Material certificate 3.1
- A 3.1 material certificate is an inspection certificate under EN 10204 in which the manufacturer confirms that the material meets the requirements of the order, with results from specific inspection, validated by an inspection representative independent of manufacturing.
- MCP (Model Context Protocol)
- MCP is an open standard for how AI models connect to external data and tools in a secure, structured way.
- Multi-tenant
- Multi-tenant is an architecture in which one installation of a piece of software serves many customers (tenants), with each customer's data kept separate.
- MVP (minimum viable product)
- An MVP (minimum viable product) is the smallest version of a product that can be put in the hands of real users to test the most important assumption behind it.
- Offline-first
- Offline-first is a design principle in which an app works without a network connection, stores what the user does locally and synchronises when the connection returns.
- Post-project costing (estimate vs actual)
- Post-project costing compares what a project or order actually cost with the estimate the quote was based on.
- Project accounting
- Project accounting is the tracking of income and costs per project: budget or estimate, hours and purchases incurred, amount invoiced and the expected result when the project is finished.
- Prompt engineering
- Prompt engineering is the craft of formulating instructions to a language model so it returns consistent, precise, and useful answers.
- RAG (Retrieval-Augmented Generation)
- RAG is a technique where a language model answers based on the business's own documents — instead of only its general training.
- RPA (Robotic Process Automation)
- RPA is software that mimics human clicking in a user interface — typically used to integrate against legacy systems without an API.
- Serial-number service history
- Serial-number service history is the complete record of everything done to one specific unit after delivery: service, repairs, replaced parts, inspections and certificates.
- Supplier Document Requirements List
- A Supplier Document Requirements List (SDRL) is the customer's list of which documents a supplier must deliver, in what format and by when, as part of the contract or purchase order.
- Vector database
- A vector database is a database optimised for storing and searching embeddings — the foundation of RAG systems and AI search.
Fixed price per phase · Full source code, owned by you · Reply within one business day
We build what you need. Let's figure out what that is.
Book a 30-minute meeting, or send us a few lines about what's grinding. We reply within one business day.