Glossary

Enterprise technology glossary

Who this is for

Plain-language definitions of the enterprise technology terms that come up when you compare vendors, from zero trust to RPO and RTO. Each entry says what the term means, why it matters when you are buying, and how it connects to related terms, with links to the matching category hub and evaluation guide.

Written by Techarda editors · Updated 22 Oct 2026. Want the news behind a term? Topics follow each subject across categories.

AI & ML Platforms

AI guardrails: AI guardrails are controls around a model or AI application that check inputs and outputs against rules, such as blocking sensitive data, off-topic requests or unsafe responses.
Embeddings: Embeddings are lists of numbers that represent the meaning of text, images or other data, so that similar items sit close together and can be found by similarity search.
Fine-tuning: Fine-tuning is further training of an existing model on a smaller, task-specific dataset so it performs better on that task or follows a particular style or format.
Inference: Inference is running a trained model to produce an output, such as a prediction, an answer or generated text.
Large language model (LLM): A large language model is an AI model trained on very large amounts of text to predict what comes next, which lets it answer questions, summarise, write and work with code.
MLOps: MLOps is the set of practices and tools for deploying, monitoring and maintaining machine learning models in production, including versioning of data, code and models.
Retrieval-augmented generation (RAG): Retrieval-augmented generation is a pattern where a system first searches your own content for passages relevant to a question, then gives them to a language model to write an answer based on those sources.

AI Agents & Automation

Agentic AI: Agentic AI describes AI systems designed to pursue goals with some autonomy, planning, acting and checking their own progress, rather than only answering a single prompt.
AI agent: An AI agent is software that uses a language model to decide on a sequence of steps and carry them out by calling tools, such as searching, updating records or running code, to reach a goal.
Human in the loop: Human in the loop is a design where a person reviews, approves or corrects an AI system’s output or action at set points before it takes effect.
Model Context Protocol (MCP): The Model Context Protocol is an open standard for connecting AI applications to tools and data sources through a common interface, so a connector written once can serve many assistants and agents.
Prompt injection: Prompt injection is an attack where text in a prompt, document, email or web page tells an AI model to ignore its instructions and do something else, such as leak data or take an unwanted action.
Robotic process automation (RPA): Robotic process automation is software that copies the clicks and keystrokes a person makes in applications to automate repetitive, rule-based tasks.

Data & Analytics

Change data capture (CDC): Change data capture is a technique that reads inserts, updates and deletes from a database’s transaction log and passes them on as events to other systems.
Data lakehouse: A data lakehouse is an architecture that stores data in open formats on low-cost object storage while adding the table management, transactions and performance of a data warehouse.
Data warehouse: A data warehouse is a database designed for analytical queries over large volumes of structured data, usually loaded from operational systems.
ETL and ELT: ETL (extract, transform, load) and ELT (extract, load, transform) are ways of moving data from source systems into an analytics platform, differing in whether data is reshaped before or after it is loaded.
Event streaming: Event streaming is the continuous capture, storage and processing of events, such as transactions, clicks or sensor readings, as they happen, usually in an ordered, durable log.
Open table format: An open table format is a public specification for organising data files on object storage as tables, with schemas, transactions and version history, so different engines can share them.

Databases

Distributed SQL database: A distributed SQL database spreads data and queries across many nodes, often in several regions, while keeping the transactions and SQL interface of a relational database.
NoSQL database: A NoSQL database stores data in models other than relational tables, such as documents, key-value pairs, wide columns or graphs, usually for flexibility or scale.
Vector database: A vector database stores embeddings and quickly finds those most similar to a query, using indexes built for nearest-neighbour search.

Cloud Infrastructure

Egress fees: Egress fees are charges a cloud or storage provider applies when data leaves its network, and sometimes when it moves between the provider’s own regions or zones.
FinOps: FinOps is an operating practice that brings engineering, finance and business teams together to manage cloud spending, so costs are visible, owned and weighed against value.
Serverless computing: Serverless computing is a cloud model where the provider runs your code or service on demand and manages the servers underneath, charging for actual use rather than reserved capacity.
Shared responsibility model: The shared responsibility model sets out which security and operational duties belong to a cloud provider and which to the customer, for each type of service.
Sovereign cloud: A sovereign cloud is a cloud service built to keep data, operations and legal control within a specific jurisdiction, usually with local operation and limits on foreign access.

Hybrid & Multicloud

Hybrid cloud: Hybrid cloud is an approach that runs workloads across on-premises infrastructure or private cloud and one or more public clouds, managed as one estate.
Hyperconverged infrastructure (HCI): Hyperconverged infrastructure combines compute, storage and networking on standard servers managed by software as one system, instead of separate storage arrays and servers.
Hypervisor: A hypervisor is software that runs several virtual machines on one physical server, sharing its processors, memory and storage between them.
Multicloud: Multicloud means using services from more than one public cloud provider, either for different workloads or for the same application.

Containers & Kubernetes

Container: A container is a lightweight package holding an application and everything it needs to run, isolated from other containers while sharing the host’s operating system kernel.
Kubernetes: Kubernetes is an open-source system for deploying, scaling and managing containers across a cluster of machines.
Kubernetes operator: A Kubernetes operator is software that runs in a cluster and automates running one application, such as a database, by encoding tasks like deployment, backup, failover and upgrades.
Service mesh: A service mesh is an infrastructure layer that manages communication between services, adding encryption, traffic control and telemetry without changing application code.

DevOps & Platform Engineering

CI/CD: CI/CD stands for continuous integration and continuous delivery or deployment: building and testing every code change automatically, then releasing it through a repeatable pipeline.
DORA metrics: DORA metrics are a small set of software delivery measures from long-running industry research, covering how often teams deploy, how long changes take to reach production, how often changes fail and how quickly service is restored.
GitOps: GitOps is an operating model where the desired state of systems is declared in a Git repository and automated agents keep running environments in line with it.
Infrastructure as code (IaC): Infrastructure as code is the practice of defining infrastructure, such as networks, servers and cloud services, in version-controlled files that tools apply automatically.
Internal developer portal: An internal developer portal is a single interface where engineers find services, owners, documentation and templates, and start common tasks such as creating a new service.
Platform engineering: Platform engineering is the discipline of building and running an internal platform that gives development teams self-service tools and paved paths for building, deploying and operating software.

Developer Tools

AI coding assistant: An AI coding assistant is a tool that uses language models to suggest, write, explain or review code inside an editor, terminal or code review workflow.
Feature flag: A feature flag is a switch in code that turns a feature on or off for some or all users at run time, without deploying new code.
Software bill of materials (SBOM): A software bill of materials is a machine-readable list of the components, libraries and versions that make up a piece of software.

Observability

AIOps: AIOps is the use of machine learning and, increasingly, language models in IT operations to group alerts, detect anomalies, suggest causes and automate routine responses.
Distributed tracing: Distributed tracing follows a single request as it passes through many services, recording timing and context at each step.
Observability: Observability is the ability to understand a system’s internal state from the data it produces, mainly metrics, logs and traces, including for problems nobody anticipated.
OpenTelemetry: OpenTelemetry is an open-source project that provides standard APIs, software development kits and a collector for producing and sending metrics, logs and traces.
Service level objective (SLO): A service level objective is a target for how reliably a service should perform from its users’ point of view, such as the share of requests that succeed within a given time.

Security

Cloud-native application protection platform (CNAPP): A CNAPP combines cloud security posture management, workload protection, permission analysis and code or image scanning in a single platform.
Extended detection and response (XDR): Extended detection and response collects and correlates security data from endpoints, identities, email, cloud and network in one system to detect, investigate and respond to threats.
Identity and access management (IAM): Identity and access management covers the systems and processes that create identities, authenticate users and services, and control what each one may access.
Passkey: A passkey is a sign-in credential based on public-key cryptography, stored on a device or in a password manager, that replaces a password and resists phishing.
Post-quantum cryptography (PQC): Post-quantum cryptography is the family of encryption and digital signature algorithms designed to stay secure against attacks by large quantum computers.
Privileged access management (PAM): Privileged access management controls and monitors accounts with powerful rights, such as administrators and service accounts, by vaulting credentials, granting time-limited access and recording sessions.
Security information and event management (SIEM): A SIEM collects logs and security events from across an organisation, stores them, and runs searches and detection rules to find threats and support investigations and audits.
Zero trust: Zero trust is a security approach that never grants access on network location alone and instead verifies identity, device and context for every request.
Zero trust network access (ZTNA): Zero trust network access gives users access to specific applications after checking identity and device, instead of placing them on the network as a VPN does.
Zero-day vulnerability: A zero-day is a software vulnerability that attackers know about or are exploiting before the vendor has released a fix.

Networking

Microsegmentation: Microsegmentation divides a network or environment into small zones, down to individual workloads, and allows only the traffic each one needs.
SD-WAN: SD-WAN is software-defined wide area networking: it connects sites over a mix of links, such as broadband, mobile and private circuits, and routes traffic by application policy from central management.
Secure access service edge (SASE): SASE is an architecture that combines wide area networking, such as SD-WAN, with cloud-delivered security services including secure web gateway, cloud access security broker and zero trust network access.

Storage & Data Protection

Immutable backup: An immutable backup is a copy of data that can’t be changed or deleted for a set retention period, even by an administrator.
Isolated recovery environment: An isolated recovery environment is a separate, tightly controlled environment that holds protected copies of critical data and systems, where they can be checked and restored after a cyber attack.
Object storage: Object storage keeps data as objects with metadata in a flat namespace, reached over web APIs, and scales to very large volumes at low cost.
RPO and RTO: Recovery point objective (RPO) is the most data you can afford to lose, measured as time since the last good copy, and recovery time objective (RTO) is the longest you can afford a service to be down.

HPC & Accelerated Compute

GPU cluster: A GPU cluster is a group of servers fitted with graphics processing units and joined by a fast network so they can work together on AI training, inference or scientific computing.
HPC job scheduler: An HPC job scheduler hands out a cluster’s compute resources to queued jobs according to priorities, quotas and resource needs.
Liquid cooling: Liquid cooling removes heat from servers by circulating liquid to or around the hottest components, instead of relying on air alone.
Quantum computing: Quantum computing uses quantum bits, which can hold combinations of states at once, to tackle certain problems that are impractical for classical computers.

Edge & IoT

Edge computing: Edge computing processes data close to where it is produced, such as in factories, shops, vehicles or telecom sites, rather than sending it all to a central cloud or data centre.
Industrial IoT (IIoT): Industrial IoT is the use of connected sensors, machines and controllers in industrial settings to collect data and monitor or control physical processes.
Operational technology (OT) security: OT security protects the systems that monitor and control physical equipment, such as industrial control systems, building management and utility infrastructure.

Enterprise Applications

Enterprise resource planning (ERP): ERP software runs an organisation’s core business processes, such as finance, procurement, inventory and manufacturing, on a shared set of data.
IT service management (ITSM): IT service management is the set of practices and tools for delivering and supporting IT services, including incident, request, change and asset management.
Low-code platform: A low-code platform lets people build applications and workflows mostly through visual tools and configuration, adding code only where needed.