Find the right part of the ecosystem.
Products, standards and concepts by function, each with its source, maturity and evidence label. Placement reflects documented function. It cannot be bought.
142 results · curated records · Browse companies →
Google DeepMind's Gemini models. The newest, Gemini 4 Argon, targets software engineering, knowledge work and cyber defense, but is so far limited to selected testers.
Genesys Cloud Agentic Virtual Agent handles customer requests across voice and digital channels and carries out multi-step tasks in back-office systems within set guardrails, configured in AI Studio.
Copilot cloud agent researches a repository, plans changes and implements them in the background in an ephemeral cloud environment, opening pull requests.
Google's open-source framework for building, debugging and deploying AI agents and multi-agent workflows, available in Python, TypeScript, Go, Java and Kotlin.
Google Cloud component that controls and secures traffic between users, agents and tools (including MCP servers), enforcing identity, access and content-safety policies.
A Google Cloud catalog where organizations register, find and govern their AI agents, MCP servers, tools and skills.
A mode in Google Search that answers questions with an AI-written response, using Gemini models, and lets users ask follow-up questions, with links to web sources.
Conversational shopping in AI Mode plus agentic checkout that buys a tracked item via Google Pay at your price, with permission.
A curated library of prebuilt agent samples and templates in Gemini Enterprise Agent Platform to help developers build agents quickly.
Google Cloud's Gemini agent for business users: it plans tasks, uses tools and company systems, and produces documents, answers, content and code within enterprise controls.
UCP payment handlers that let shoppers pay during agent checkout: Google Pay with Google Wallet on Gemini and AI Mode, and Shop Pay on UCP-compatible platforms.
A shopping cart that works across stores in Google Search and the Gemini app (YouTube and Gmail planned), tracks prices and stock, and checks out with Google Pay or on the merchant's site.
OpenAI's API model catalog; current flagships GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna, across capability and cost tiers.
Guidewire framework, launched in its Qusar release (Aug 2026), for insurers to build, deploy and manage AI agents with access to policy, claims and billing data.
Scheduling, authorization, clinical admin; high regulatory and access constraints.
Liability insurance for small and medium US businesses covering lawsuits from their use of AI, sold through partner insurers' business policies. Announced March 2026, pending regulatory approval.
RFC 9110 (HTTP Semantics, IETF Internet Standard) defines the architecture, terminology and semantics shared by all HTTP versions.
HubSpot's AI agents, including customer, prospecting and data agents, plus a builder for custom agents, run on HubSpot CRM data and are managed in Agent Hub.
Agent that moves commercial insurance broker submissions from intake to a decision-ready file within the carrier's pricing, appetite and authority rules, with underwriter review where needed.
Third-party measurement of how AI assistants and agents make choices: which options they find, which they finally recommend, why, and what that means for sales.
Instinct is a personal assistant you text or call that handles life admin such as bills, gifts, groceries, appointments and trips.
Real-time car insurance quote comparison inside ChatGPT, live in 11 US states with a rollout to 40+ states planned for 2026. Choosing, advice and buying the policy remain separate steps.
Fin is an AI customer agent built on proprietary models that resolves customer issues across service, sales and success channels.
Remote MCP server that lets outside AI agents call Intercom's Fin as a tool to answer support questions and run Fin procedures; requires a Fin Agent API key.
Klaviyo AI agents for online retailers: Composer drafts marketing flows and campaigns from a prompt, and Customer Agent answers shoppers' questions using the store's customer data.
LangChain's MIT-licensed open-source framework for building and running agents with low-level control over workflows, memory, streaming and human-in-the-loop checks.
LangSmith is LangChain's agent and LLM observability platform, giving teams visibility into what their agents are doing.
AI systems that help lawyers and legal teams with legal research, document review, drafting and related legal workflows.
AI agent platform that automates work across connected business apps, acting as an AI teammate for repetitive tasks.
An open protocol that lets software agents pay for APIs and services over HTTP, using the 402 'payment required' response.
Lets AI agents get insurance quotes and hand the purchase back to the insurer with the required disclosures. Production use is Marrow's own claim. Marrow also stewards the AMI draft standard.
An AI-assisted platform for Marsh's brokers in the London insurance market that helps match and send placement requests, negotiate terms and bind contracts. Brokers approve every placement.
Mastercard's agentic payments program using Agentic Tokens. Announced Apr 2025; enabled for US cardholders by late 2025; live test transactions with 17 Latin American issuers and processors (Mar 2026).
Model Context Protocol: the standard way an AI agent connects to tools and data. Created by Anthropic (Nov 2024); donated to the Linux Foundation's Agentic AI Foundation on 9 Dec 2025.
An official MCP extension that lets tools return interactive interfaces, such as forms, charts and dashboards, shown inside the chat; supported by ChatGPT, Claude, Goose and VS Code.
The merchant destination agents reach through storefronts, commerce APIs and business-access protocols.
Meta Model API is Meta's inference platform for Muse Spark and other Muse models, compatible with OpenAI and Anthropic SDKs.
Muse is Meta's personal AI agent that works across a person's apps from a dedicated cloud computer. Rolling out in the US on iOS, Android and web.
Muse Code is Meta's terminal and CI coding agent that plans, edits and runs commands, with approvals and an OS sandbox.
Muse Spark is Meta's hosted model for agentic and coding work. Version 1.3 is recommended for new work, with a 1M-token context window.
Agent registry, identity, management, observability and governance.
Microsoft's consumer AI assistant. Copilot Checkout lets people buy inside Copilot, using the Agentic Commerce Protocol with PayPal, Shopify and Stripe (rolling out in the US from Jan 2026).
Agent Store is the central hub in Microsoft Copilot for discovering and installing agents from Microsoft, partners and your organization.
Copilot Studio is a platform to build, customize, deploy and manage AI agents and workflows with built-in security and governance.
Lets a Copilot Studio agent connect to an external agent that supports the Agent2Agent (A2A) protocol and hand it tasks during a conversation or workflow.
Preview feature that exposes a published Copilot Studio agent as an MCP server or A2A agent so outside apps and agents can call it, in early-release environments only.
Insurance that covers losses when an AI system underperforms: one way the risk of agent errors can be transferred.
Connected marketing, commerce and business workflow agent with approvals.