Enterprise AI stack guide
Every layer of the enterprise AI stack, compared with sources.
From the cloud platforms up to the agents your staff use: 101 tools across building, controlling and using AI, each claim linked to a primary source. Updated weekly.
The enterprise AI stack
Teams build agents, control them, and put AI in people's hands, all on top of a cloud platform. Pick a layer or a category to compare the tools in it.
Use
AI that employees and developers work with
Control plane
Govern, secure, and observe every agent
- Registry & management9 tools
- Identity & access10 tools
- Guardrails & security10 tools
- Monitoring & evaluation7 tools
- AI & MCP gateways8 tools
Build & orchestrate
Write agents and chain them into workflows
Foundation
Cloud AI platforms the stack runs on
Start with your question
Not sure which category you need? Pick the job you are trying to get done.
Roll out Copilot, ChatGPT or Claude to staff safely
Pick the assistant, then put data protection, ownership and access rules around it before the first wave of users.
4 stepsPut our first agent into production
Build it with a framework your team can support, wire it into the process, then route, watch and guard it before go-live.
5 stepsGet control of the agents we already have
Find every agent first, give each one an owner and identity, then apply policy and monitoring across the fleet.
4 stepsDecide whether to allow always-on agents
These agents act without a prompt each time, so treat the decision like granting a new employee access, not like installing an app.
4 steps
What changed recently
All updates- 2026-10-09 · DeprecationGemini Code Assist
Google stopped selling new Gemini Code Assist Standard and Enterprise subscriptions on October 9, 2026. Existing subscriptions auto-renew through 2026, auto-renewal ends in 2027, and Google points customers to Antigravity.
Source - 2026-10-09 · ReleaseNanoClaw
NanoClaw v2026.10.0 moved to calendar versioning with stable, beta and edge update channels. OneCLI installs must upgrade the gateway to 1.42.0 to close a credential-injection bypass, and a WhatsApp message-spoofing flaw was patched.
Source - 2026-10-07 · ReleaseMLflow
MLflow 3.17.0 adds typed judge decisions, fine-grained permissions for shared resources and faster SQL trace analytics. It requires a coordinated database upgrade.
Source
More ways to decide
Every claim links to a primary source, no listing or ranking here can be bought, and the whole dataset is open.
Nothing here can be bought: no listing fees, sponsored placement or paid badges.
Frequently asked questions
What does enterpriseai.tools track?
Microsoft Foundry, Amazon Bedrock, and the Gemini Enterprise Agent Platform as the cloud foundation layer, plus the leading tools to build, control and use AI agents that enterprises pair with them — each tracked record carries a verifiable source URL.
How often is the data updated?
A weekly source-backed scan refreshes every tracked record. Each per-tool page shows a 'last reviewed' date alongside its governance posture, and the /updates feed surfaces newly verified releases, license changes, certifications, and acquisitions.
Is there a signup, paywall, or data capture?
No. The whole site runs as a static export with no accounts, no email capture, and no advertising. Google Analytics is the only third-party script, it loads only if you accept the consent prompt, and declining means no request to Google at all. The 'Help me evaluate' flow runs entirely in your browser either way.
Where do the underlying facts come from?
Vendor trust centers, compliance documentation, repository LICENSE files, and primary product docs. Every governance claim on a per-tool page links its source URL directly so claims can be audited.