Agentic AI Glossary

An open-source programming framework for conversational agents, multi-agent networks, tools, human collaboration, structured output, and evaluation.

An AWS platform for building, deploying, operating, securing, connecting, remembering, observing, and evaluating agents using different frameworks and models.

The earlier managed Bedrock agent service for orchestrating foundation models, action groups, knowledge bases, and user requests.

Managed agent infrastructure in the Claude platform for deploying and operating agents without owning every runtime component directly.

A Python and TypeScript SDK that exposes the agent loop, tools, and context-management capabilities used by Claude Code.

An agentic coding tool that reads codebases, edits files, runs commands, and works through development tasks in terminal, IDE, desktop, and browser surfaces.

An agent framework centered on collaborative crews, role-based agents, tasks, tools, memory, knowledge, guardrails, and production flows.

A structured event-driven workflow layer in CrewAI for controlling state, execution order, routing, persistence, and agent participation.

Databricks capabilities for authoring, deploying, tracing, evaluating, and operating custom agents and GenAI applications on governed enterprise data.

An open-source framework for production-ready RAG, search, multimodal pipelines, tools, memory stores, evaluations, and iterative agents.

A cloud software-development agent that researches repositories, plans changes, edits code on a branch, and can open pull requests for human review.

GitLab's AI-native platform for agentic chat, foundational and custom agents, flows, sessions, tools, and asynchronous work across the software lifecycle.

Google's open-source framework for building, debugging, evaluating, and deploying agents and multi-agent systems across models and tools.

The managed execution layer within Google's agent platform for running, scaling, isolating, and operating deployed agents and supporting services.

Google Cloud's unified platform for building, deploying, governing, integrating, and optimizing enterprise-grade agents and model-based applications.

A compact open-source Python library for building agents with code-generation or structured tool-calling execution styles.

IBM's enterprise platform for building, cataloging, connecting, orchestrating, deploying, and using agents, tools, applications, and MCP servers.

A batteries-included agent harness with planning, subagents, filesystem tools, context management, and long-running execution on LangGraph.

A configurable open-source framework that provides model, tool, prompt, middleware, retrieval, and agent-loop abstractions across many integrations.

A low-level orchestration runtime for long-running, stateful agents with durable execution, streaming, persistence, interrupts, and human input.

A platform for tracing, evaluating, testing, monitoring, deploying, and improving language-model and agent applications across frameworks.

A data and agent framework focused on connecting language-model applications to documents, indices, retrieval systems, tools, workflows, and memory.

An open-source Microsoft-originated framework for conversational agents, multi-agent patterns, tools, and event-driven agent applications.

A managed Microsoft knowledge layer that connects enterprise data into reusable, permission-aware knowledge bases for agents.

Microsoft's multi-language framework for building agents and explicit multi-agent workflows with state, middleware, telemetry, checkpoints, and human input.

A graphical low-code studio for building and managing agents, workflows, agent flows, tools, knowledge connections, and publishing channels.

A managed Microsoft Foundry service for building, hosting, deploying, observing, and scaling agents across supported frameworks and models.

An open-source Microsoft SDK for integrating models, prompts, plugins, memory, and agent patterns into .NET, Python, and Java applications.

An open platform for tracing, evaluating, monitoring, reviewing, and improving GenAI applications, RAG systems, and agents across development and production.

A framework-agnostic toolkit for connecting, profiling, observing, evaluating, and optimizing enterprise agent workflows across existing agent frameworks.

A lightweight software development kit for building tool-using agents, handoffs, guardrails, sessions, and traced agent runs.

A software-engineering agent that can inspect repositories, plan work, edit files, run commands, and complete coding tasks across several interfaces.

OpenAI's unified API surface for model responses, tool use, conversations, streaming, and stateful application interactions.

Oracle's design-time environment for creating, configuring, validating, and deploying agents and multi-agent workflows inside Fusion Applications.

A Python agent framework built around typed dependencies, validated outputs, model portability, tools, graphs, evaluation, and observability.

Salesforce's platform for creating, customizing, testing, governing, and deploying business agents across Salesforce data, workflows, and channels.

ServiceNow's central orchestration component for planning, selecting, coordinating, and supervising agents that collaborate on complex platform workflows.

ServiceNow's managed authoring environment for configuring, testing, and deploying AI agents, tools, knowledge, triggers, and agentic workflows.

A fully managed Snowflake platform for agents that plan tasks, call tools, execute code, and reason across governed structured and unstructured data.