Resources
AI & Data Glossary
Cutting through the jargon: clear, professional definitions of the concepts behind modern AI, machine learning, and data systems — grouped so you can find what you need fast.
Foundations
The core building blocks behind every AI and data system.
- Artificial Intelligence (AI)
- The broad field of building systems that perform tasks normally requiring human intelligence — recognizing patterns, making predictions, understanding language, and making decisions.See our AI strategy work →
- Machine Learning (ML)
- A subset of AI where systems learn patterns from data rather than following hand-written rules, improving their predictions as they see more examples.See our ML services →
- Deep Learning
- A branch of machine learning using multi-layered neural networks, particularly effective for unstructured data like images, audio, and text.See our ML services →
- Neural Network
- A model loosely inspired by the brain, made of layered nodes that transform input data through weighted connections to learn complex patterns.
- Big Data
- Datasets too large, fast-moving, or varied for traditional tools to process — typically requiring distributed storage and processing systems to extract value from.See our data engineering work →
- Algorithm
- A defined sequence of steps a computer follows to solve a problem or complete a task — the underlying logic behind any AI or software system.
- Model
- The output of a training process: a mathematical representation that has learned patterns from data and can make predictions or generate content on new inputs.
- Training Data
- The examples a model learns from. Quality, volume, and relevance of training data are usually the biggest drivers of how well a model performs.
Generative AI & Language Models
The technology behind chatbots, copilots, and content-generating systems.
- Generative AI
- AI systems that create new content — text, images, code, audio — rather than only classifying or predicting from existing data.See our Generative AI services →
- Large Language Model (LLM)
- A deep learning model trained on massive amounts of text, able to understand and generate human-like language across a wide range of tasks.See our Generative AI services →
- Transformer
- The neural network architecture behind modern LLMs. Its "attention" mechanism lets a model weigh the relevance of every word in context at once, rather than reading strictly left to right.See our Generative AI services →
- Foundation Model
- A large, general-purpose model (like GPT or Claude) pre-trained on broad data, designed to be adapted to many specific tasks rather than built from scratch each time.
- Prompt Engineering
- The practice of designing inputs to an AI model to reliably get accurate, useful, and well-formatted outputs — a key skill in building production LLM applications.
- Fine-Tuning
- Further training a pre-trained model on a narrower, task-specific dataset so it performs better on a particular use case or adopts a specific tone or format.
- Tokenization
- The process of breaking text into smaller units (tokens) that a language model processes — roughly word-fragments — used both to read input and generate output.
- Embedding
- A numerical representation of text, images, or other data that captures meaning in a way computers can compare — the basis for semantic search and RAG.
- Hallucination
- When an AI model generates confident-sounding but false or fabricated information — one of the key risks that production LLM systems need to guard against.
Retrieval & Knowledge Systems
How AI systems stay grounded in real, private, and up-to-date data.
- RAG (Retrieval-Augmented Generation)
- A technique that retrieves relevant information from your own documents or database before an LLM generates a response — grounding answers in real data and reducing hallucination.See our RAG implementations →
- Vector Database
- A database optimized to store and search embeddings, enabling fast "semantic" search — finding results by meaning rather than exact keyword match.See our data engineering work →
- Semantic Search
- Search based on the meaning of a query rather than exact keywords, powered by embeddings — returning relevant results even when wording differs.
- Knowledge Base
- A structured collection of an organization's information — documents, FAQs, policies — that AI systems like copilots and chatbots draw on to answer questions accurately.
AI Agents & Automation
Systems that don't just answer questions — they take action.
- AI Agent
- An AI system that can plan, use tools, and take multi-step actions toward a goal — rather than just responding to a single prompt with a single answer.See our AI agent work →
- Agentic AI
- An approach to building AI systems that operate with a degree of autonomy — reasoning through multi-step tasks, making decisions, and adapting their plan as they go, with varying levels of human oversight.See our automation services →
- Multi-Agent System
- An architecture where multiple specialized AI agents collaborate — each handling a sub-task — coordinated to complete a larger, more complex workflow.See our automation services →
- Tool Use / Function Calling
- The ability of an AI model to call external tools, APIs, or functions — searching the web, querying a database, sending an email — as part of completing a task.
- Orchestration
- The logic that coordinates multiple AI calls, tools, and agents into a reliable workflow — managing order, retries, and handoffs between steps.
Operationalizing AI
Running AI systems reliably in production, not just in a demo.
- MLOps
- The practices and tooling for deploying, monitoring, and maintaining machine learning models in production — the ML equivalent of DevOps.See our ML services →
- LLMOps
- MLOps practices adapted specifically for large language model applications — covering prompt versioning, evaluation, cost monitoring, and safety guardrails.See our Generative AI services →
- AIOps
- Using AI to automate and improve IT operations itself — such as detecting anomalies, predicting outages, or triaging incidents faster than manual monitoring.See our cloud & DevOps services →
- Model Monitoring
- Ongoing tracking of a deployed model's accuracy, latency, and behavior in production — catching problems before they impact users or the business.See our managed support →
- Model Drift
- The gradual decline in a model's accuracy over time as real-world data shifts away from what it was originally trained on — a key reason production models need monitoring.
- CI/CD for ML
- Continuous integration and deployment pipelines adapted for machine learning — automatically testing, validating, and rolling out new model versions safely.See our cloud & DevOps services →
Data & Infrastructure
The pipelines and platforms that feed every AI system.
- Data Pipeline
- An automated sequence of steps that moves and transforms data from its source to where it's usable — for reporting, analytics, or feeding a model.See our data engineering work →
- ETL / ELT
- Extract, Transform, Load (or Extract, Load, Transform) — the standard patterns for moving data from source systems into a warehouse or lake, in a clean, usable structure.See our data engineering work →
- Data Warehouse
- A central repository built for structured, query-friendly storage of business data — the backbone of most reporting and BI dashboards.See our data engineering work →
- Data Lake
- A storage system that holds large volumes of raw data in its native format — structured or not — until it's needed, offering more flexibility than a warehouse.
- Data Governance
- The policies and processes that ensure data is accurate, secure, and used appropriately — increasingly critical as AI systems rely on more sensitive data.See our security & compliance services →
Responsible & Applied AI
Where AI meets real business problems — and real-world accountability.
- Computer Vision
- AI systems that interpret and understand visual data — detecting objects, reading documents, or inspecting products from images and video.See our ML services →
- Natural Language Processing (NLP)
- The field of AI focused on understanding and generating human language — powering everything from sentiment analysis to chatbots and document processing.See our ML services →
- Predictive Analytics
- Using historical data and statistical models to forecast future outcomes — demand, churn, risk — so businesses can act ahead of time rather than react.See our ML services →
- Recommendation System
- A model that predicts what a user is likely to want next — products, content, actions — based on their behavior and similarity to other users.See our ML services →
- Explainable AI (XAI)
- Techniques that make an AI model's decisions understandable to humans — increasingly important for trust, debugging, and regulatory compliance.See our AI strategy work →
- AI Bias & Ethics
- The practice of identifying and mitigating unfair or harmful patterns an AI system may have learned from its training data, and ensuring responsible use overall.See our AI strategy work →
Ready to put these concepts to work?
Whichever part of the stack you're thinking about — strategy, a generative AI feature, or the data pipeline behind it — we can help you scope it properly.