Artificial Intelligence (AI) Glossary

As Artificial Intelligence (AI) and automation become commonplace in the modern enterprise, it’s key for IT professionals at all levels to understand the terms used to describe these tactics. Below are some of the most-used terms alongside short, easy to understand definitions.

• AI-augmentation: Partnership of people and AI designed to enhance human intelligence and decision making. Sometimes shortened to augmentation.

• AIOps: Artificial intelligence for IT operations. A best practice that allows organizations to reduce noise, triage faster with the right context, and scale event-driven automation.

• Automation: The use of technology to perform tasks with minimal human intervention, including event-driven automation, human-in-the-middle automation, IT process automation, and business workflow automation.

• Benchmarking: Comparing the performance of systems, processes, or products against established standards or competitors to identify areas for improvement.

• Bias: The inaccuracy of a decision made by an AI system when it has learned from unrepresentative data or a flawed model.

• Business process automation: The use of technology to automate repeatable, day-to-day tasks. This is also known as BPA.

• Business process management: The systematic management of business processes to achieve organizational goals and optimize performance.

• Chatbot: A computer program designed to simulate conversation with human users, typically through text or voice interactions.

• ChatGPT: An instance of OpenAI’s GPT (Generative Pre-trained Transformer) model specifically fine-tuned for generating conversational responses.

• CI/CD pipeline: Continuous Integration/Continuous Deployment pipeline, a set of automated processes for building, testing, and deploying software changes.

• Cognitive computing: Computer simulation of human thought processes, often involving artificial intelligence techniques. It can be used synonymously with AI in certain scenarios.

• Configuration drift: The gradual and unintended divergence of system configurations from their intended state.

• Configuration management: The process of establishing and maintaining the consistency of a system’s performance and functional attributes.

• Conversational AI: Artificial intelligence technology that enables computers to engage in natural language conversations with users.

• Corpus: A large and structured set of texts used for linguistic analysis, training language models, or building datasets.

• Data augmentation: Techniques used to increase the size or diversity of a dataset by creating variations of existing data points.

• Data drift: Refers to the difference between the data an AI model encounters in the real world compared to the data it was trained on.

• Data lake: A centralized repository that allows you to store structured, semi-structured, and unstructured data at scale.

• Data mining: The process of discovering patterns, correlations, or insights from large datasets using computational techniques.

• Data validation: The process of ensuring that data meets quality and integrity standards before it is used for analysis or decision-making.

• DataOps: The practice of leveraging software and data engineering, quality assurance, and infrastructure operations into a single organization.

• Deep learning: A subset of ML based on an artificial neural network with 3 or more layers.

• DevSecOps: The integration of security practices into the DevOps process to ensure that security is built into software development and deployment workflows.

• Diffusion: A method of ML that takes an existing piece of data, like a video or photo, and adds noise to the background. Diffusion models can then train AI to recover the initial data.

• Explainability: The degree to which the decisions or outputs of an AI system can be understood and interpreted by humans.

• Extensibility: The ability of a system or platform to be easily extended or customized with additional functionality.

• Extraction: The process of identifying and retrieving relevant information or features from data sources.

• Fine tuning: The process of adjusting the parameters or architecture of a machine learning model to improve its performance on a specific task or dataset.

• Foundational model: A pre-trained machine learning model that serves as the basis for further customization or specialization such as OpenAI’s GPT-4.

• Generative AI: A type of AI deep learning model that can generate text, audio, images, code, and other content based on training data. This is also known as GenAI.

• GPT: Generative Pre-trained Transformer, a large language model developed by OpenAI.

• Grounding: The process of connecting abstract concepts or symbols to real-world objects or experiences.

• Hallucination: A response that AI generates containing false or misleading information presented as fact.

• Human-in-the-loop: The need for human interaction, intervention, and judgment to control a process or to help train machine learning models; related to AI-augmentation.

• Hyperautomation: The automation of business processes by introducing AI, ML, and RPA.

• Intelligent process automation: The combination of artificial intelligence and automation technologies to streamline and optimize business processes.

• Interpretability: The ability to understand and explain how a machine learning model arrives at its decisions or predictions.

• IT automation: Automating repetitive tasks and processes within IT infrastructure to improve efficiency and reliability.

• IT process automation: Automating IT-related workflows and processes to reduce manual effort and improve consistency.

• Large language model: A type of AI that uses deep learning techniques and large corpuses of data to generate original content. This is also known as LLM.

• Low code/no code: Development approaches that enable the creation of software applications with minimal or no traditional coding.

• Machine learning: A subfield of AI that allows a machine to learn complex relationships between data to improve performance on tasks without specific instruction. It is also known as ML.

• Microservice: A piece of a software architecture pattern where an application is composed of small, loosely coupled services that communicate over a network.

• Model: A mathematical representation or algorithm used to power predictions, classifications, or decisions based on input data.

• Multimodal language model: A type of language model capable of understanding and generating text, images, audio, or other forms of data.

• Natural language processing: A field of artificial intelligence focused on enabling computers to understand, interpret, and generate human language.

• Neural network: A machine learning framework inspired by the structure and function of the human brain, composed of interconnected nodes (neurons) organized in layers.

• Orchestration: The coordination and management of automated tasks, processes, or services to achieve a desired outcome.

• Parameter: A variable or setting that helps describe the behavior or configuration of a system, model, or algorithm.

• Playbooks: Prescriptive guides or documents that outline step-by-step procedures for carrying out specific tasks or processes.

• Predictive analytics: A form of business intelligence that applies ML to generate a predictive model for certain business applications.

• Process automation: Automating repetitive tasks and workflows within a business process to improve efficiency and reduce errors.

• Prompt engineering: The process of designing and refining prompts or input stimuli to elicit desired responses from AI systems.

• Provisioning: The process of allocating and configuring resources, such as servers or software, to support the operation of IT systems or applications.

• Reinforcement learning: A type of machine learning where an agent learns by interacting with its environment and receiving rewards or penalties based on its actions.

• Robotic process automation: A form of business process automation that allows users to create scripts to automate digital tasks.

• Role-based access control: A security model that restricts system access based on the roles and responsibilities of individual users within an organization.

• Rule-based system: A system or software application that operates according to a set of predefined rules or conditions.

• Self-managed systems: Systems or processes that can autonomously monitor, adjust, and optimize their own performance without human intervention.

• Structured data: Data that is organized into a specific format, such as tables or databases, with clearly defined fields and relationships.

• Supervised learning: A type of machine learning where the model is trained on labeled data, with input-output pairs provided during training.

• Synthetic data: Artificially generated data that mimics the real-world data characteristics and is used for training and testing machine learning models.

• Task automation: Automating specific tasks or activities to improve efficiency and productivity.

• Transformer: A type of neural network architecture commonly used in natural language processing tasks, known for its ability to handle long-range dependencies.

• Trigger: An event or condition that initiates or activates a specific action or process.

• Unstructured data: Data that does not have a predefined format or organization, such as text documents, images, or sensor data.

• Unsupervised learning: A type of machine learning where the model learns patterns or structures in data without explicit instruction or labeled examples.

• Up-skilling: Training and adapting employees’ skills to work more effectively with automation and AI technologies. It is also known as reskilling.

• Virtual assistant: An AI-powered software application that provides assistance or performs tasks for users through natural language interactions.

• Workflow: A series of interconnected tasks or activities that are organized and executed to achieve a specific goal.

• Workflow automation: The automation of repetitive or manual tasks within a workflow to improve efficiency and consistency.

If you’re eager to learn more about AI and automation and turn this glossary into real-life practical learnings, read the Deconstructing AI and Automation eBook.