DEPLOYMENT

Two ways to run it. You choose.

The product is identical. What changes is where it lives and who holds the keys.

OPTION 01

SaaS

Fully managed, hosted by us. OpenParser runs on Innovatix's secure cloud infrastructure. Fastest to get started — sign up and load documents the same day.

OPTION 02

Private cloud

For organizations that need full control over where their data lives, OpenParser can be deployed on your own infrastructure — local servers, your company data center, or your private cloud environment (AWS, Azure, or similar). Your documents stay within infrastructure you control.

How your data is handled

Your documents stay in a secure, access-controlled environment - hosted on our infrastructure, or entirely within yours if you choose on-premises deployment.
Your documents are never used to train any model - ours or anyone else's. What you upload stays yours.
Answers come only from your own sources - OpenParser doesn't fill gaps with general AI knowledge, so you're never getting a plausible-sounding guess.
Only the people you authorize can see it - access follows your organization's own roles, so sensitive documents stay visible only to the right people.
You control what goes in - documents are added to your knowledge base deliberately, not pulled in automatically from anywhere they happen to sit.
Deploy where your policy requires - cloud, private cloud, or fully on-premises, so data residency isn't a blocker to adoption.

Access & accountability

Everyone works from the same knowledge base - with the right boundaries. Your whole team gets consistent answers, while each person only sees what their role allows.
Access follows your organization's structure. Permissions are set by who someone is and what they're authorized to see, not left open by default.
A user's answers only draw on what they're allowed to see. Even when a question touches a restricted document, the response respects that person's access level.
Administrators stay in control. Your team manages who has access and at what level, so the knowledge base grows without losing oversight.
Nothing is shared outside the boundaries you set. Knowledge stays inside your organization's access rules — it doesn't leak across teams or departments that shouldn't see it.
Accountability is built in, not bolted on. Access control is part of how OpenParser works, not an add-on feature you have to configure separately.

MODEL GOVERNANCE

Your documents, your choice of model.

Most AI tools decide which model processes your documents. OpenParser doesn’t.

Llama by default

Included and ready to run — including fully local, with no external API call.

Your commercial model

OpenAI, Anthropic or Gemini, using your own contracted keys and terms.

Your private endpoint

Self-hosted or internally governed models via a private endpoint.

Separated contracts

The model your organization approves is the model that sees your documents.

Built for organizations that cannot afford a guess.