OnVeris

PRIVATE AI DEPLOYMENT

Private AI deployment on infrastructure you control

Run AI models on your own servers or in your private cloud, and manage users, assistants and company knowledge from one place. Your IT team deploys it; staff use it through chat and the applications they already work in.

YOUR PRIVATE AI DEPLOYMENT

Staff and applicationsOnVeris Chat · OpenAI-compatible API
OnVeris GatewayUsers · Assistants · Knowledge · Access
Model NodesLocal inference on your hardware
Company knowledgeApproved document collections

Choose where your AI runs

With a private deployment, model inference happens inside an environment you manage. You decide where requests are processed, which models are available, and how staff and applications reach them. OnVeris runs on-premises or in your private cloud. The Gateway and model Nodes can live on one machine or be split across several, depending on how you want to scale.

On-premises

Install OnVeris inside your own network and run models on hardware you manage. You control the machines, network access and the local services the deployment depends on - a fit for organisations that want prompts and internal knowledge processed within their existing infrastructure.

Private cloud

Deploy inside private cloud infrastructure that you manage, or that a provider manages for you. You pick the environment - processing capacity and network connections included - and keep the OnVeris installation and model choices in your own hands.

Manage the deployment through one Gateway

OnVeris keeps the administration of your AI platform separate from the hardware running its models. The Gateway handles access and routes requests to connected Nodes. Start with both on one machine, and add Nodes as capacity or model needs grow.

  • Gateway - users, assistants, knowledge bases, API keys and connectors, managed from one interface.
  • Nodes - each runs locally hosted models on its own hardware. Nodes can offer different models, or simply add capacity.
  • Chat and applications - staff use OnVeris Chat; supported applications connect through the OpenAI-compatible API.
OnVeris Gateway monitoring interface.
OnVeris Node configuration interface.

Choose models and hardware around the work

Start from the tasks you want AI to handle. Answering from a policy collection, drafting a response and working through a long technical document each place different demands on a model - model size, context length, GPU memory and the number of concurrent requests all move performance. CPU inference works, but for production workloads we recommend GPU acceleration. Before you commit to hardware, run representative questions and documents through it and check both quality and speed. Models can be assigned to particular assistants, with fallback models where appropriate. Gateway monitoring and analytics then show you model activity, response times, failures and demand as usage grows.

Put private AI in front of staff

Staff use OnVeris Chat in a browser, on desktop or on mobile. An administrator builds assistants for particular tasks - model, instructions, and the people who get access. For applications, the OpenAI-compatible API takes configured API keys, so developers can connect supported software to your deployment and make specific assistants available inside existing workflows.

Bring company knowledge into the deployment

Build knowledge bases from policies, product manuals and internal procedures. A locally hosted embedding model indexes the content, and assistants can pull in relevant passages when answering. Knowledge access is configured alongside assistant sharing, so you decide which collections each assistant can use and who can use that assistant.

Governed AI Knowledge Access →

Start with one useful deployment

The first rollout is a chance to assess the environment and the user experience with real work. A good starting point: one defined assistant, one document collection, and the group of staff who'll use it.

  1. Choose the environment. Decide where the Gateway and Nodes will run, who manages them, and how users connect.
  2. Test a model against real tasks. Run representative work through it to check hardware capacity, answer quality and response times.
  3. Configure the assistant and its access. Set the instructions, knowledge and tools, then share it with the intended users.
  4. Review, then extend. Feedback and Gateway analytics will tell you when it's time to add knowledge, assistants, models or capacity.

Plan your private AI deployment with Parker Software

Tell us where you want AI to run, what it should do for your team, and which information it should use. We'll demonstrate the platform and talk through the infrastructure and configuration your deployment would need.

Private AI deployment FAQs

Which operating systems does OnVeris support?

The OnVeris installer supports Windows, Linux and macOS. The Gateway and Node can run together on one machine or on separate machines. Linux servers can also use the command line installer.

What hardware do we need for a private AI deployment?

There's no single answer - the model, context length and number of simultaneous requests set the memory and processing you need. CPU inference is available, and a suitable GPU improves performance. Evaluate your chosen models with representative workloads before sizing a wider rollout.

Can we use locally hosted models without per-token inference charges?

Locally hosted model use has no per-token inference fees from OnVeris. Budget for your OnVeris licence, infrastructure and operation of the deployment. Costs for any external services you connect depend on those services.

What happens when we connect a cloud business system?

A configured connector communicates with the relevant service using its supported authentication. Those service requests follow the connection you configure, while your locally hosted models continue to run on your OnVeris Nodes.