Skip to content
Enterprise

intoCHAT Enterprise: SSO, custom roles, audit log and custom limits

What intoCHAT's Enterprise plan adds for larger teams: SAML single sign-on, custom roles, a 365-day audit log, access to selected agents, custom domains, custom limits, and the API, MCP server and helpdesk integrations around it.

Author
intoCHAT Team
Published
Last updated
Reading time
8 min read
Illustration: an enterprise team with role badges, an audit review and system connections around an AI chat agent
On this page
  1. What Enterprise adds to Pro
  2. Sign-in and access
  3. A record of every change
  4. Control over each agent
  5. Channels and models
  6. Integrations: API, MCP and helpdesks
  7. Data and privacy
  8. A rollout plan that works
  9. Talk to us about Enterprise
  10. Frequently asked questions

intoCHAT Enterprise is the plan for organizations where many people work on AI agents, security reviews are part of buying software, and volume is too high for a fixed monthly allowance. It runs on the same product as every other plan, and adds the controls a larger team needs: sign-in through your identity provider, roles you define, a long audit trail and limits agreed with you.

This guide lists what the plan includes today, how each control works, and what intoCHAT does not offer, so your review can rely on it.

What Enterprise adds to Pro

Pro ($49/month)Enterprise (custom)
Agents, messages, training characters, crawls10 agents, 10,000 messages/month, 1M characters per agent, 20 crawls/monthAgreed with you
Team members5 (extra seats $9/month)Unlimited
Single sign-on (SAML 2.0)NoYes
RolesAdmin, Editor, ViewerAlso custom roles
Access to selected agentsYesYes
Audit log90 days365 days
Custom domainOne per agentOne per agent
Procedures per agent10Up to 50
Contacts / helpdesk tickets per agent25,000 / 1,000100,000 / 5,000
AI modelsEvery modelEvery model
SupportPriority supportDedicated support and an SLA

Every plan has the full agent builder. Plan details and limits are in Plans and limits.

Sign-in and access

Single sign-on

On Enterprise, your team signs in with your identity provider instead of a separate intoCHAT password. Single sign-on works with Okta, Microsoft Entra ID, Google Workspace and any provider that supports SAML 2.0.

  • You verify your email domains with a DNS record. Only addresses on those domains can use SSO, and a domain can belong to one account only.
  • Someone who signs in with SSO for the first time joins your team automatically, with the role you choose. Nobody needs an invitation.
  • With Require SSO on, password and Google sign-in are refused for those addresses, and multi-factor rules stay with your identity provider. The account owner is never locked out.
  • Not included: SCIM provisioning, roles from IdP groups, sign-in started at the IdP, OpenID Connect and more than one identity provider per account.

Roles and custom roles

Every plan with seats gives each person Admin, Editor or Viewer. On Enterprise, the owner can also define up to 20 custom roles, each with None, View, Edit or Manage for every area of the account: agent settings, knowledge, actions and procedures, leads and contacts, conversations and live chat, analytics, integrations and webhooks, team, billing and API keys. A support lead can manage conversations without touching integrations, and a content editor can edit knowledge without seeing leads.

Access to selected agents

From Starter, you can limit a member to Selected agents. They then see only those agents everywhere, from the dashboard to transcripts, leads, exports and the live chat inbox, and other agents return "not found". This suits agencies, separate brands or regional teams in one account. Each person can also turn on two-factor sign-in with an authenticator app.

A record of every change

The audit log under Settings, Audit log shows who did what and when: agents created, changed or deleted, knowledge sources, actions and procedures, webhooks and integrations, API keys and connected apps, team and role changes, plan changes and downloaded exports. You can filter by action, member, agent and date, and export up to 10,000 rows as CSV.

It never records passwords, API keys, tokens or other secrets, message content or visitor details, and IP addresses are shortened to their network. Pro keeps 90 days; Enterprise keeps 365.

For agent settings there is also a settings history: every saved version of an agent, who saved it and what changed, with a one-click restore.

Control over each agent

Each agent has its own controls on the Share tab, on every plan:

  • Allowed domains and Private agent, so the agent only answers on your sites, or nowhere while it isn't ready.
  • Limits and access: messages per visitor per hour, a monthly message cap for the agent, and blocked countries.
  • Identity verification: your server signs the signed-in user's id, so actions can look up that user's own data, and Only verified visitors can chat keeps an agent for logged-in customers. See secure chatbot for logged-in users.
  • Team approval (Pro and Enterprise): an API action or a procedure step waits until someone from your team approves it. See human approval for AI actions.

A custom domain (Pro and Enterprise) puts an agent's Help Page, chat link and widget script on a subdomain of yours, such as chat.yourcompany.com, with a CNAME record.

Channels and models

The same agent answers in a website widget, on a hosted Help Page, in Slack and by email, with replies from your own domain on Pro and Enterprise. It also answers on WhatsApp, Facebook Messenger and Instagram, with up to 20 channels per agent on Enterprise; your team can pause the AI and reply from the transcript. Voice and SMS are not offered.

Enterprise includes every AI model, chosen per agent: Claude Haiku 5.5 by default, GPT-6 Luna, Gemini Flash, Claude Sonnet 5.5, Gemini Pro, Claude Opus 5.5 or Auto.

Integrations: API, MCP and helpdesks

  • REST API with an OpenAPI spec: chat with your agents, read conversations, leads and knowledge sources, sync contacts and manage webhooks. API keys can be limited to the scopes they need, with up to 20 active keys per account.
  • MCP server: connect Claude.ai, ChatGPT, Claude Code, Cursor and other AI tools to your account, read-only if you choose. The person's role and agents still apply.
  • Signed webhooks for leads, hand-offs and live chat requests, form submissions, bookings, return requests and orders.
  • Helpdesk tickets: each hand-off can open a ticket in Zendesk, Freshdesk, HubSpot, Gorgias or Help Scout, with an AI summary, the transcript and the fields you map.
  • Knowledge connectors: Google Drive, Notion, solved Zendesk and Freshdesk tickets and Intercom Help Center articles, on top of website crawls, files and Q&A pairs.
  • Custom API actions, up to 25 per agent, with bearer token, basic auth, API key or custom header authentication, and procedures that chain them step by step.

Enterprise also covers custom integrations agreed with our team. intoCHAT has a Zapier app. There is no Make app yet; webhooks and the REST API work with Make and n8n.

Data and privacy

  • According to the privacy policy, the database is hosted in Switzerland. Some providers process data in the United States and other countries, under standard contractual clauses where the law requires them.
  • To write a reply, the visitor's message, recent history and the relevant parts of your knowledge go to the provider of the model you chose: Anthropic by default, or OpenAI or Google. intoCHAT doesn't train AI models on your data or your visitors' data.
  • Credentials for connected services are encrypted and never shown again, and webhooks are signed.
  • Conversations are kept until you delete them. You can let visitors delete their own chat and offer temporary chats that are never saved.
  • A data processing agreement is available on request at privacy@intochat.ai. Our chatbot GDPR guide covers what you need to do on your side.
  • intoCHAT is a cloud service only, with no on-premises installation, and holds no SOC 2, ISO 27001 or HIPAA certification.

A rollout plan that works

  1. Pilot one agent on Pro. Train it on your help content, connect your helpdesk and test it with simulated customers.
  2. Set up access. Decide who needs Admin, which teams need a custom role, and who works only on selected agents.
  3. Connect your identity provider and verify your domains, then switch on Require SSO.
  4. Lock down each agent: allowed domains, limits, identity verification and approvals for sensitive actions.
  5. Review the audit log and transcripts weekly during the first month, and add Q&A pairs for weak answers.

Talk to us about Enterprise

Tell us your expected volume, your identity provider and the systems your agents should reach, and we'll put together an Enterprise plan. For the full list of controls, see enterprise AI chatbot security.

Contact Sales View Pricing

SSO • Custom roles • 365-day audit log • Custom limits

Frequently asked questions

What does the intoCHAT Enterprise plan include?

Everything in Pro, plus single sign-on with SAML 2.0, custom roles, an audit log that keeps 365 days, unlimited team members and limits for agents, messages, knowledge and crawls agreed with you. Enterprise also includes every AI model, up to 50 procedures per agent, 100,000 contacts and 5,000 helpdesk tickets per agent, custom integrations, dedicated support and an SLA. The price depends on your volume.

Does intoCHAT support SSO and SCIM?

Single sign-on, yes: on Enterprise your team signs in through Okta, Microsoft Entra ID, Google Workspace or any SAML 2.0 identity provider, joins the team automatically with the role you choose, and can be required to use SSO. SCIM provisioning is not available, so removing someone at your identity provider doesn't remove them from your intoCHAT team; remove them under Settings, Team as well.

Is intoCHAT SOC 2 or ISO 27001 certified?

No. intoCHAT holds no security or data protection certification such as SOC 2, ISO 27001 or HIPAA. It offers SAML single sign-on, custom roles, an audit log, two-factor sign-in, encrypted credentials and signed webhooks, and a data processing agreement on request. According to the privacy policy, the database is hosted in Switzerland. If your policy requires a certification, intoCHAT doesn't meet it today.

How does intoCHAT connect to our internal systems?

Several ways. Custom API actions let the agent call your own HTTP API during a chat, up to 25 per agent. The REST API with an OpenAPI spec and the MCP server let your systems and AI tools such as Claude and ChatGPT read conversations, leads and knowledge. Signed webhooks send leads, hand-offs, forms and bookings to your server. Hand-offs can open tickets in Zendesk, Freshdesk, HubSpot, Gorgias or Help Scout, and the agent can learn from Google Drive, Notion, solved Zendesk and Freshdesk tickets and Intercom articles.

Which AI models can we use on Enterprise?

Every model intoCHAT offers, chosen per agent: Claude Haiku 5.5 (the default), GPT-6 Luna and Gemini Flash at 1 message per reply, Claude Sonnet 5.5 and Gemini Pro at 2, Claude Opus 5.5 at 5, or Auto. Messages go to the provider of the model you choose, and intoCHAT doesn't train AI models on your data.

How can we test intoCHAT before committing to Enterprise?

Start on the free plan or a 7-day trial of Starter, build an agent from your own content and test it in the Playground, with tests that simulate customers. Pro already includes the audit log (90 days), a custom domain, team approval for actions and routing, so you can check most of the setup before you sign. Then contact sales with your expected volume.

Back to Blog
  • Illustration of five chat assistants facing a smartphone with a customer message, a clipboard with a bar chart beside them and the Alps behind
    Study · Switzerland

    Do AI Chatbots Understand Swiss German? We Tested 5 Models

    50 customer messages from 7 regions, 5 current AI models, 1,000 answers: every model understood written Swiss German almost as well as standard German. The differences are in how they reply.

    6 min read

  • Editorial illustration of a closed shop inside a browser window, with a warm chat light connecting it to a visitor on a laptop at night
    Lead generation

    The 11:47 PM Customer: Your Best Lead Is Still Awake

    Someone is on your pricing page tonight, one unanswered question away from leaving. Here is how to build an AI lead generation chatbot that helps them take the next step—with a reusable prompt, ten tests and a seven-day launch plan.

    10 min read