One assistant, one job
Build assistants around specific tasks and teams. An HR assistant that answers policy questions, a support assistant that pulls up customer or order details, a writing assistant that sticks to your house style - each one does one thing well.
FROM CONFIGURATION TO CONVERSATION
A focused assistant: A familiar place to use it
You decide how each one works
Write the instructions, pick the model, choose the knowledge bases it can draw on. From there, staff can pick the assistant in OnVeris Chat, and your applications can call it through the OnVeris API.
Only the tools it needs
You choose which approved connector tools each assistant gets. In Chat, you can also have selected tools ask the user first, so a person reviews the action before it happens.
Access, in your hands
You control who can use each assistant. An admin can build one for a particular role or task and grant access to the right users or service keys - no one else.
Keep AI where it belongs
Run OnVeris on-premises or in a private cloud you control, with locally hosted models doing the work. Your team gets task-specific AI without prompts leaving for an external model API.
No per-token fees for locally hosted model use.
From chat to purposeful AI
Chat is where everyday conversations happen. Search makes approved documents available to AI. Assistants define how AI handles specific tasks. Connectors link the approved systems behind it all.
Put task-specific AI to work
See how OnVeris assistants could support your teams while staying within the lines you set.