IRIS ENTERPRISE
Keep Your Company IP In-House With Self-Hosted AI
Keep Your Company IP In-House With Self-Hosted AI
Put AI agents to work on your internal documents, research, and business processes. We build the agents and configure self-hosted models to process proprietary information inside your company’s environment.
HOW IT WORKS
Iris Turns Company Goals Into Coordinated Execution
Custom agents collect company context, understand what needs to happen next, and execute across your tools to keep operations moving.
HOW IT WORKS
Iris Turns Company Goals Into Coordinated Execution
Custom agents collect company context, understand what needs to happen next, and execute across your tools to keep operations moving.
HOW IT WORKS
Iris Turns Company Goals Into Coordinated Execution
Custom agents collect company context, understand what needs to happen next, and execute across your tools to keep operations moving.
WORK ACROSS YOUR TOOLS
Run Workflows Across the Tools Your Team Already Uses
Define which systems agents can access, which actions are allowed, and where information may be sent. Connections are reviewed as part of the private deployment scope.
Book a demo
Book a demo
BUILT FOR BUSINESS RESULTS
Control Where Company Information Is Processed
Define which documents agents can use, where models process them, and what gets retained. We scope the deployment around those rules, including connected tools and operating logs.
Company Memory
Key decisions, customer context, and operating knowledge stay available across teams and workflows.
Custom Skills
Your processes, tool usage, and approval rules become repeatable skills agents can execute.
PRIVATE COMPANY WORKFLOWS
Put Agents to Work on Proprietary Information
Internal research and document workflows
Build agents that search internal documents, compare information, and prepare reports from approved sources. Scope the processing environment and review rules before connecting confidential material.
INTERNAL DOCUMENT RESEARCH
Find answers across authorized documents and bring the relevant sources into a reviewable brief.
CONTROLLED MODEL ACCESS
Define which supported model endpoints may process each class of information. External processing is included only where it is explicitly allowed.
REVIEWABLE DATA FLOWS
Document the path through tools, models, memory, logs, and backups so your team can assess the full deployment.
Scope the Agent and Model Environment Together
Private cloud or on-premises
Review agent hosting, model endpoints, company memory, integrations, and logs as one deployment. The implementation plan identifies where each part runs and which external services, if any, are involved.
Requirements for local inference, restricted data access, retention, and action approvals are agreed before the build.
PRIVATE COMPANY WORKFLOWS
Put Agents to Work on Proprietary Information
Internal research and document workflows
Build agents that search internal documents, compare information, and prepare reports from approved sources. Scope the processing environment and review rules before connecting confidential material.
Scope the Agent and Model Environment Together
Private cloud or on-premises
Review agent hosting, model endpoints, company memory, integrations, and logs as one deployment. The implementation plan identifies where each part runs and which external services, if any, are involved.
REVIEWABLE DATA FLOWS
Document the path through tools, models, memory, logs, and backups so your team can assess the full deployment.
CONTROLLED MODEL ACCESS
Define which supported model endpoints may process each class of information. External processing is included only where it is explicitly allowed.
INTERNAL DOCUMENT RESEARCH
Find answers across authorized documents and bring the relevant sources into a reviewable brief.
Requirements for local inference, restricted data access, retention, and action approvals are agreed before the build.
PRIVATE COMPANY WORKFLOWS
Put Agents to Work on Proprietary Information
Internal research and document workflows
Build agents that search internal documents, compare information, and prepare reports from approved sources. Scope the processing environment and review rules before connecting confidential material.
CONTROLLED MODEL ACCESS
Define which supported model endpoints may process each class of information. External processing is included only where it is explicitly allowed.
Scope the Agent and Model Environment Together
Private cloud or on-premises
Review agent hosting, model endpoints, company memory, integrations, and logs as one deployment. The implementation plan identifies where each part runs and which external services, if any, are involved.
INTERNAL DOCUMENT RESEARCH
Find answers across authorized documents and bring the relevant sources into a reviewable brief.
REVIEWABLE DATA FLOWS
Document the path through tools, models, memory, logs, and backups so your team can assess the full deployment.
Requirements for local inference, restricted data access, retention, and action approvals are agreed before the build.
QUESTIONS ABOUT IRIS
Everything You Need to Know Before Deploying Iris
How Iris integrates with your systems, runs across teams, handles company data, and adapts to your operating environment.
Tell us which workflow you want to improve. We’ll help define the implementation.
Talk to Us
Talk to Us
Can Iris Be Integrated Into Our Enterprise Environment?
Yes. Iris can be deployed in different ways depending on your infrastructure and data requirements. For standard deployments, Iris can connect to approved business systems and operate across the tools your teams already use. For companies that need more control, Iris Enterprise is designed to support a customer-controlled private cloud or a separately scoped on-premises deployment. Enterprise deployments can also include SSO, role-based access, private model endpoints, policy-based model routing, scoped credentials, action approvals, and audit records.
Can Iris Be Integrated Into Our Enterprise Environment?
Yes. Iris can be deployed in different ways depending on your infrastructure and data requirements. For standard deployments, Iris can connect to approved business systems and operate across the tools your teams already use. For companies that need more control, Iris Enterprise is designed to support a customer-controlled private cloud or a separately scoped on-premises deployment. Enterprise deployments can also include SSO, role-based access, private model endpoints, policy-based model routing, scoped credentials, action approvals, and audit records.
How Does Iris Integrate With Our Existing Tools?
Iris works across the systems already used to run your company, including CRM, email, documents, project management tools, internal databases, and other business applications. Agents can retrieve information from those systems, combine it into the context needed for a workflow, and take approved actions such as preparing updates, creating tasks, drafting follow-ups, or updating records. Standard connectors support defined actions, while custom integrations can extend Iris into internal software and specialized workflows.
How Does Iris Integrate With Our Existing Tools?
Iris works across the systems already used to run your company, including CRM, email, documents, project management tools, internal databases, and other business applications. Agents can retrieve information from those systems, combine it into the context needed for a workflow, and take approved actions such as preparing updates, creating tasks, drafting follow-ups, or updating records. Standard connectors support defined actions, while custom integrations can extend Iris into internal software and specialized workflows.
What Work Can Iris Handle?
Iris is designed for recurring operational work where information has to be collected, understood, coordinated, and turned into action. That can include preparing account and meeting briefs, researching companies and markets, monitoring customer renewals, identifying blockers across projects, preparing CRM updates, drafting follow-ups, coordinating hiring processes, producing recurring reports, investigating engineering issues, and tracking commitments and next steps. Different Iris applications are designed around Operations, Sales, Customer Success, Research, Recruiting, Executive Assistance, and Engineering.
What Work Can Iris Handle?
Iris is designed for recurring operational work where information has to be collected, understood, coordinated, and turned into action. That can include preparing account and meeting briefs, researching companies and markets, monitoring customer renewals, identifying blockers across projects, preparing CRM updates, drafting follow-ups, coordinating hiring processes, producing recurring reports, investigating engineering issues, and tracking commitments and next steps. Different Iris applications are designed around Operations, Sales, Customer Success, Research, Recruiting, Executive Assistance, and Engineering.
How Are Iris Agents Customized for Different Departments?
Each department can have agents built around the workflows that matter to that team. A Sales agent might understand the company’s account-review process, use CRM and email, prepare call briefs, and suggest record updates. An Operations agent might track project commitments, identify blockers, and prepare follow-ups. Customization includes workflow steps, connected systems, information access, output format, approval requirements, and permitted actions.
How Are Iris Agents Customized for Different Departments?
Each department can have agents built around the workflows that matter to that team. A Sales agent might understand the company’s account-review process, use CRM and email, prepare call briefs, and suggest record updates. An Operations agent might track project commitments, identify blockers, and prepare follow-ups. Customization includes workflow steps, connected systems, information access, output format, approval requirements, and permitted actions.
How Does Company Memory Work?
Iris can maintain useful company context across workflows instead of treating every request like a new conversation. That can include account history, previous decisions, operating playbooks, project context, saved preferences, and past workflow results. When relevant and authorized, that information can be used in future work. Access still follows the company’s permissions, so shared memory does not mean every employee or agent automatically gains access to every piece of company information.
How Does Company Memory Work?
Iris can maintain useful company context across workflows instead of treating every request like a new conversation. That can include account history, previous decisions, operating playbooks, project context, saved preferences, and past workflow results. When relevant and authorized, that information can be used in future work. Access still follows the company’s permissions, so shared memory does not mean every employee or agent automatically gains access to every piece of company information.
Can Different Teams Use Iris Together?
Yes. Iris can support workflows that move across departments while keeping access scoped to the people and systems involved. For example, a customer renewal can involve Sales, Customer Success, Legal, Finance, and Engineering. Iris can carry authorized customer context and outstanding commitments into that workflow, identify what is blocking the outcome, and coordinate the next actions. Each person still operates within their permissions.
Can Different Teams Use Iris Together?
Yes. Iris can support workflows that move across departments while keeping access scoped to the people and systems involved. For example, a customer renewal can involve Sales, Customer Success, Legal, Finance, and Engineering. Iris can carry authorized customer context and outstanding commitments into that workflow, identify what is blocking the outcome, and coordinate the next actions. Each person still operates within their permissions.
Which AI Models Can Iris Use?
Iris is designed to be model-agnostic. Different supported models can be used depending on the task, required capability, cost, speed, and the company’s data policies. For Enterprise deployments, model selection can also depend on where the information is allowed to be processed. The company’s workflow, memory, permissions, and integrations do not need to be built around one model provider.
Which AI Models Can Iris Use?
Iris is designed to be model-agnostic. Different supported models can be used depending on the task, required capability, cost, speed, and the company’s data policies. For Enterprise deployments, model selection can also depend on where the information is allowed to be processed. The company’s workflow, memory, permissions, and integrations do not need to be built around one model provider.
Plan a Private AI Deployment for Your Company
Book a demo to review the workflow, infrastructure, model requirements, and data controls your implementation needs. We’ll define the deployment scope before the build.