Glossary
Technical concepts we work with every day, written for the people we work with. No jargon walls. No assumed knowledge.
30 terms
AI/ML
Agent orchestration is the coordination layer that manages how AI agents are invoked, sequenced, and monitored within a workflow. It handles task routing, state management, error recovery, and human escalation - so agents work together reliably at production scale.
Read full definitionAI/ML
Agentic AI refers to AI systems that can plan, make decisions, and take actions autonomously to achieve a goal. Unlike simple chatbots that respond to one prompt at a time, agentic systems break complex tasks into steps, use tools, and self-correct along the way.
Read full definitionAI/ML
An AI agent is a software system that uses a large language model to plan, reason, and take actions autonomously. Unlike chatbots that respond to single prompts, agents execute multi-step workflows - calling APIs, querying databases, and making decisions to achieve a defined goal.
Read full definitionAI/ML
Computer vision is the field of AI that enables machines to interpret and act on visual information - images, video, and documents. Applications range from quality inspection in manufacturing to document processing in healthcare.
Read full definitionAI/ML
Embeddings are numerical representations of data (text, images, audio) in a high-dimensional space where similar items are located near each other. They allow AI systems to measure similarity, search by meaning, and cluster related content.
Read full definitionAI/ML
Fine-tuning is the process of training a pre-trained AI model on a smaller, domain-specific dataset to adapt its behavior for a particular task. It modifies the model's internal weights so it performs better on your specific use case without training from scratch.
Read full definitionAI/ML
AI hallucination is when a language model generates information that sounds plausible but is factually incorrect or entirely fabricated. It is a fundamental behavior of LLMs, not a bug, and managing it is essential for any production AI application.
Read full definitionAI/ML
Human-in-the-loop (HITL) is a design pattern where AI agents pause for human review, approval, or correction before executing high-stakes actions. HITL checkpoints prevent errors in critical workflows - like clinical decisions, financial transactions, or legal document reviews - while allowing agents to operate autonomously on low-risk tasks.
Read full definitionAI/ML
Inference is the process of using a trained AI model to generate predictions or outputs from new inputs. When you send a prompt to an LLM and get a response, that is inference. It is where compute costs, latency, and user experience are determined.
Read full definitionAI/ML
A large language model is a neural network trained on massive text datasets to understand and generate human language. LLMs power chatbots, content generation, code assistants, and most modern AI products.
Read full definitionAI/ML
MLOps (Machine Learning Operations) is the set of practices for deploying, monitoring, and maintaining machine learning models in production. It applies DevOps principles to ML systems, keeping models accurate, reliable, and cost-effective after launch.
Read full definitionAI/ML
Model Context Protocol (MCP) is an open standard developed by Anthropic that defines how AI models connect to external data sources and tools. MCP provides a universal interface for AI agents to access databases, APIs, file systems, and other services through a consistent protocol instead of custom integrations for each tool.
Read full definitionAI/ML
A multi-agent system (MAS) coordinates two or more AI agents to accomplish tasks that are too complex for a single agent. Agents in a MAS communicate, delegate subtasks, and share context - following patterns like hierarchical management, pipeline assembly lines, or peer-to-peer collaboration.
Read full definitionAI/ML
Natural language processing is the branch of AI focused on enabling machines to understand, interpret, and generate human language. It covers everything from sentiment analysis and text classification to machine translation and conversational AI.
Read full definitionAI/ML
Prompt engineering is the practice of crafting and optimizing the instructions given to a language model to get consistent, high-quality outputs. It is the most accessible and cost-effective way to improve AI application behavior without modifying the underlying model.
Read full definitionAI/ML
Retrieval-augmented generation is a technique that combines a language model with a searchable knowledge base. Instead of relying solely on what the model learned during training, RAG retrieves relevant documents first, then generates answers grounded in that specific data.
Read full definitionAI/ML
A token is the basic unit of text that a language model processes. Words, parts of words, and punctuation are all broken into tokens. Token counts determine model costs, context window limits, and response length constraints.
Read full definitionAI/ML
Tool use (also called function calling) is the ability of an AI agent to interact with external systems - APIs, databases, code execution environments, or web services. The LLM decides which tool to call, generates the required parameters, and incorporates the tool's output into its reasoning chain.
Read full definitionAI/ML
The transformer is the neural network architecture behind virtually all modern language models. Introduced in 2017, it uses a mechanism called self-attention to process entire sequences of text in parallel, making it far more efficient and capable than previous approaches.
Read full definitionAI/ML
A vector database is a specialized database designed to store and search high-dimensional numerical representations (embeddings) of data. It enables fast similarity search, which is the foundation of AI-powered search, recommendation systems, and RAG pipelines.
Read full definitionDevelopment
An API gateway is a server that acts as the single entry point for all client requests to your backend services. It handles routing, authentication, rate limiting, and request transformation so individual services do not have to.
Read full definitionDevelopment
Feature flags are toggles in your code that control which features are visible or active for which users. They decouple deployment from release, letting you ship code to production without exposing it to everyone immediately.
Read full definitionDevelopment
GraphQL is a query language for APIs that lets the client specify exactly which data it needs. Instead of multiple REST endpoints returning fixed data shapes, a single GraphQL endpoint returns precisely the fields the client requests.
Read full definitionDevelopment
Microservices is an architecture pattern where a software application is built as a collection of small, independent services that communicate over APIs. Each service handles a specific business capability and can be developed, deployed, and scaled independently.
Read full definitionDevelopment
React Server Components (RSC) are a React feature that lets components render on the server and send only the resulting HTML to the client. This reduces JavaScript bundle sizes, improves page load speed, and simplifies data fetching.
Read full definitionDevelopment
Technical debt is the accumulated cost of shortcuts, quick fixes, and suboptimal design decisions in a codebase. Like financial debt, it compounds over time - making every future change slower, riskier, and more expensive.
Read full definitionDevelopment
WebSocket is a communication protocol that enables persistent, two-way communication between a client and server over a single connection. Unlike HTTP where the client must initiate every exchange, WebSocket allows the server to push data to the client in real time.
Read full definitionBusiness
An AI readiness assessment evaluates an organization's data quality, infrastructure, team skills, and process maturity to determine where AI can deliver measurable value and what gaps need to be filled first.
Read full definitionBusiness
A proof of concept is a small-scale implementation that validates whether a technical approach or product idea works before committing full resources. It answers 'can we build this?' with working code, not slide decks.
Read full definitionBusiness
Technical due diligence is a systematic review of a company's technology assets, architecture, code quality, team capabilities, and technical risks. Investors and acquirers use it to assess whether the technology can support the business plan.
Read full definitionBuild With Us
100+ products shipped across AI, SaaS, mobile, and automation. Tell us what you are building.
Book a strategy call