Volume Assistant
An admin copilot that proposes. People who approve
Configuring an operations platform is the expensive part: someone has to translate a process into objects, fields, flows, reports and templates. The Volume Assistant does the translation. It discovers what already exists in your instance, drafts the change as a reviewable set and applies it through the same services and permissions a human admin would use
What you can do
Configure the whole platform by describing what you need
The Volume Assistant is an in-app copilot for the people who run Volume. Ask for a new object, a workflow, a report, a dashboard, an import, a template, a form or an action, and it drafts the change against your real data model. Nothing is written until a person reviews the change and approves it
Data model
Create an object with its fields, relationships, default record view, navigation tab and role grants as one ordered set. Rename, retype or retire fields, seeing what else is affected and how many records are involved before you decide.
Workflows
Draft new FlowState workflows as inactive drafts, refine existing ones with a step-by-step comparison and run safe test runs to check the result. Publishing stays a human action.
Reports and dashboards
Build reports from your fields and relationships, including advanced queries for analysts. Assemble dashboards with widgets, tabs and filters, then open them pre-filled in the builder.
Imports
Share a sample spreadsheet and get a mapped import profile with transforms and validations. Cell values are only read with your explicit consent, and only masked.
Communications, forms and actions
Email and SMS templates with their merge tags checked, forms as unpublished drafts, and record or bulk actions bound to a workflow and a form.
Triage configuration
Register objects for triage, define types, classifications, evidence rules and display layouts, and ask how a particular item was triaged.
Knows your instance
It works from your data model, catalogues and counts, sees only what you are allowed to see, and tells you where to do something in the UI when it cannot do it for you.
Governed by design
Bound to the permissions of the person asking, budgeted and metered per customer and per user, every turn and every applied change in the audit trail, and a kill switch that removes it entirely.
How it works
Three steps, no tickets to engineering
1
Describe
Ask in plain English from the panel on any page. The assistant discovers the relevant objects, fields and existing configuration before it drafts anything.
2
Review
The proposal arrives as a change set: each item with a before-and-after view, its risk grade, the permissions it needs and what else it touches. Reports, dashboards and imports open pre-filled in their builders.
3
Approve and apply
One click for additive changes, a typed confirmation for destructive ones. Changes apply under your permissions and land in the audit trail. Roll back within 72 hours.
What makes it different
Most AI assistants either answer questions or act without asking. The Volume Assistant has no way to write to your instance directly: it can only propose. It works from your data model and counts, never record values, so your customers’ data is not sent to a model. And it applies changes by the same route as a human admin, so validation, permissions and audit are identical.
Stronger together
Works With
Dashboards & Reporting
Relationship-aware reports and dashboards that filter each other.
Explore further
See the Volume Assistant on your kind of work
Bring a process you wish worked better. We will show you how it would be modelled, automated and reported on in Volume, with your objects and your language
Previous
Next
