Glossaire

Termes et concepts importants du développement logiciel, de la gestion de projet et des domaines connexes

Configuration file in the project root for environment-specific variables and sensitive data; typically not versioned.

AI systems that plan and execute multi-step tasks using tools under defined permissions.

Computer systems that perform tasks associated with human cognition such as classification, prediction and language understanding.

Business Process Model and Notation standard for diagramming business workflows.

Parallel line of development in a version-control repository.

Structured kickoff clarifying goals, constraints, stakeholders and success criteria for a project.

Formal customer or internal objection about quantity, quality, price or service deviations.

Recorded snapshot of changes in a version-control history with message and author metadata.

Structured comparison of alternative products, vendors or service approaches in a market.

AI field that derives information from images or video, such as detecting or classifying materials.

Numeric estimate of how certain a model is about a prediction or extracted field.

Lightweight isolated runtime package that includes an application and its dependencies.

Keeping software in a releasable state so approved versions can be deployed on demand.

Practice where every change that passes automated checks is deployed to production automatically.

Practice of frequently merging code and automatically building and testing each change.

Comma-separated values file format for exchanging tabular data between systems.

Self-service web application where customers access documents, status and selected master-data functions.

User-interface colour scheme with dark backgrounds and light foreground text.

Structured representation of entities, attributes and relationships stored by an application.

Fitness of data for use, covering completeness, accuracy, consistency, timeliness and uniqueness.

Versioned script that changes database schema in a controlled, repeatable way.

Culture and tooling that integrate software development and IT operations for faster, safer delivery.

Structured digital record of a product’s composition, origin and sustainability-relevant attributes.

Digital representation of a physical asset, yard or process used for monitoring and simulation.

Platform for packaging applications and dependencies into portable containers.

Design process model with discover/define and develop/deliver phases for problem and solution exploration.

Running AI inference on local devices near the data source instead of only in the cloud.

Process modelling notation linking events and functions, often used in ERP-oriented analysis.

Large body of work that is broken down into multiple user stories or tasks.

Enterprise Resource Planning software integrating core business processes on shared master and transactional data.

Reusable software foundation that provides structure and common components for applications.

Design pattern where people review, correct or approve AI suggestions before critical actions.

Networked sensors and devices that stream operational measurements into software systems.

Laravel starter kit providing authentication, teams/profile scaffolding and related application boilerplate.

Graph of entities and relations used to represent structured domain knowledge.

PHP web application framework with expressive syntax for MVC applications and common enterprise features.

Neural model trained on large text corpora to understand and generate natural language.

Laravel library for building dynamic interfaces mainly in PHP without heavy custom JavaScript.

Subset of AI where models learn patterns from historical data instead of being fully hand-coded.

Relatively stable reference data such as customers, materials, sites and contracts used across transactions.

Process gap where information is retyped or transferred manually between systems or paper and digital forms.

Smallest product increment that delivers enough value to learn from real users.

Practices for deploying, monitoring and maintaining machine-learning systems in production.

Visual draft of screens used to validate layout and interaction before full implementation.

Continuous observation of application health, performance and errors in running systems.

Machine learning that jointly uses multiple data types such as text, images and tables.

NLP task that locates and classifies entities such as names, dates, quantities and locations in text.

AI techniques for understanding and extracting structure from human language text.

Technology that converts text in images or PDFs into machine-readable characters.

Splitting large result sets into pages for efficient listing in user interfaces and APIs.

Automated sequence of build, test and deployment steps triggered by code changes.

Data analysis that estimates likely future outcomes from historical patterns.

Maintenance approach that predicts equipment risk to schedule service before failure.

Analytics that recommends actions under constraints, beyond merely predicting outcomes.

Technique that discovers and analyses real process flows from event logs.

Ordered list of desired product work items managed by the product owner.

Two-dimensional barcode that encodes data for fast scanning with phones or industrial scanners.

Stored collection of versioned project files and history, usually hosted in Git.

UI approach that adapts layouts to different screen sizes and devices.

HTTP-based application interface using resource-oriented endpoints and standard methods.

Architecture that retrieves relevant documents and conditions an LLM answer on those sources.

Security model that grants permissions according to assigned roles rather than individuals alone.

Agile framework that delivers work in time-boxed sprints with defined roles and ceremonies.

Script that inserts initial or demo data into a database after migrations.

Capability for users to complete tasks digitally without waiting for staff mediation.

Combining multiple sensor signals to obtain a more reliable estimate than any single source.

Fixed-length iteration in which a Scrum team commits to delivering a potentially shippable increment.

Strategic review of strengths, weaknesses, opportunities and threats.

Concrete unit of work required to implement a story or technical change.

Isolating data and configuration so one customer or organisation cannot access another’s records.

Open-source OCR engine commonly used in prototypes and document pipelines.

Predicting future values of sequentially ordered measurements over time.

Splitting text into tokens (words, subwords or characters) for NLP model processing.

Event-level business records created by day-to-day operations such as orders, weighings and invoices.

Short requirement statement describing a user need and the value of fulfilling it.

Storing text embeddings in a searchable index for semantic similarity retrieval.

Optimisation problem of assigning stops to vehicles under capacity and time constraints.

System for tracking changes to source code and enabling collaboration over time.

Human or camera-supported check of material quality and composition at intake.

Weighted or capacity-aware vehicle routing variant used in constrained logistics planning.

Operational list of vehicles or loads currently expected or present on the recycling yard.

Family of real-time object detection models widely used in computer-vision pipelines.

Vous avez encore des questions ?

Si vous n'avez pas trouvé le terme que vous cherchiez, n'hésitez pas à nous contacter.

Contactez-nous

Glossaire

Vous avez encore des questions ?

Options d'accessibilité

Assistant PSZ

Glossaire

.env (Environment file)

Agentic AI

Artificial Intelligence (AI)

BPMN

Branch

Briefing

Claim / Complaint

Commit

Competitive Analysis

Computer Vision

Confidence Score

Container

Continuous Delivery (CD)

Continuous Deployment

Continuous Integration (CI)

CSV

Customer Portal

Dark Mode

Data Model

Data Quality

Database Migration

DevOps

Digital Product Passport

Digital Twin

Docker

Double Diamond

Edge AI

EPC (Event-driven Process Chain)

Epic

ERP System

Framework

Human-in-the-Loop

Internet of Things (IoT)

Jetstream

Knowledge Graph

Laravel

Large Language Model (LLM)

Livewire

Machine Learning (ML)

Master Data

Media Break

Minimum Viable Product (MVP)

MLOps

Mockup

Monitoring

Multimodal Learning

Named Entity Recognition (NER)

Natural Language Processing (NLP)

Optical Character Recognition (OCR)

Pagination

Pipeline

Predictive Analytics

Predictive Maintenance

Prescriptive Analytics

Process Mining

Product Backlog

QR Code

Repository

Responsive Design

REST API

Retrieval-Augmented Generation (RAG)

Role-Based Access Control (RBAC)

Scrum

Seeder

Self-Service

Sensor Fusion

Sprint

SWOT Analysis

Task

Tenant Separation

Tesseract

Time-Series Forecasting

Tokenization

Transactional Data

User Story

Vector Indexing

Vehicle Routing Problem (VRP)

Version Control

Visual Inspection