The AI agent that works your operations.
A model-agnostic AI agent that reads your business data and takes action on it, across the products you already run.
Watching your workspace
Order #4821 is missing a signed document.
Booking #2091 has had no reply for six days.
The problem
Your operations run on tribal knowledge and twelve tabs.
Teams hold the work together across inbox, spreadsheet and product. The context and the busywork both cost hours, on every job.
The work is scattered.
Orders, bookings, quotes and dispatch live in different systems, and the context that connects them lives in someone's head.
The knowledge walks out the door.
When a senior team member leaves, the answers they memorised leave with them. Onboarding the next one starts the hunt again.
A generic chatbot cannot touch your records.
Confident prose with no tenant, no guardrails and no audit trail is not something operations can run on.
It knows the work
Reads the business, not a textbook.
Enzo works from the records your team already keeps, so every answer and every action fits the actual job.
Orders and jobs
What is open, what is late, and what it depends on.
Bookings and schedules
Dates, resources and the slots that need attention.
Quotes and estimates
Pricing, scope and what is still unanswered.
Customers and contacts
Who to follow up with, and the history that matters.
Invoices and documents
What is signed, what is missing, and what is overdue.
Dispatch and staffing
Crews, routes and the jobs that need a nudge.
It does the work
Reading is the start. Enzo changes the record.
Enzo is model-agnostic: it routes each task across models and acts through your tenant, your autonomy level and the audit log. Call it from any assistant over MCP.
- Create tasks
- Add notes
- Log exceptions
- Update status
- Draft quotes
- Follow up
- Flag risk
- Reconcile records
- Nudge dispatch
Money and outbound stay at Suggest by default. Quoting and customer messages keep a human in the loop until you say otherwise.
How an action travels
One agent, many industries.
Enzo is vertical-agnostic. The same agent reads the records and takes the actions each business needs, configured per product. Moving and relocation is one; restaurants and hospitality is another.
Moving and relocation
Operations platform
Reads shipments, orders, estimates, leads and dispatch, then creates tasks, adds notes, logs exceptions and updates status in the workspace.
Restaurants and hospitality
app.runtaba.com
Reads orders, bookings and customer records, then takes the same kinds of actions: tasks, notes, exceptions and status updates.
Runs across more Dumont products
Enzo is in use beyond these two. The capability is the same everywhere: the records and the verbs are configured for each product, with the same autonomy ladder and audit trail.
Wherever your team works.
The same agent, reachable from the tools people already have open. No new destination to learn.
Call it over MCP
Claude, ChatGPT and any MCP-capable client reach Enzo at askenzo.ai/mcp and act in your tenant. Enzo routes across models, so it is never tied to one.
Ask from the field
Teams message Enzo on WhatsApp and get answers or actions back where the work happens.
Embedded in the product you run
Enzo sits inside Moveezi, Taba and other Dumont products, next to the record it is working on.
Grounded in
-
US CBP import regulationsCustoms and Border Protection rules.
-
US DoD DP3 personal propertyPublic material on personal property moves.
-
National customs authoritiesPublic government sources by country.
-
Your own documentsThe files your team uploads, kept in your tenant.
How deep one capability can go: customs.
Customs and document answers grounded in public government sources and your own documents. It is one example of Enzo going deep in a domain, and every answer shows where it came from.
Instant, sourced answers
Ask in plain language. Enzo retrieves from public sources and your documents, and returns an answer you can act on, in seconds.
Cites the source, exports the proof
Every answer names the source it came from, and exports to PDF for the file.
Your documents included
Enzo grounds answers in the public sources and the files your team uploads, and keeps them inside your tenant.
You decide how much Enzo is trusted.
Autonomy is set per capability, per tenant. Nothing runs past the level you chose, and every action is on the record.
Watch
Predicts. Never acts.
Enzo flags what is about to go wrong and stays out of the way. You get the warning, not the change.
Suggest
Proposes. A human approves.
Enzo drafts the action and waits. Nothing changes until someone on your team approves it.
Run
Acts within guardrails.
Enzo performs the action, fully audited and reversible. A breach demotes Run back to Suggest, never a silent block.
This is the answer to "I cannot trust AI in my ops". Start everyone at Watch, earn Run capability by capability, on your own data.
Built for operations
Grounded
Answers cite the source they came from. No source, no answer.
Audited
Every action, approval and reversal is written to the action log.
Multi-tenant
Your data stays in your tenant. It is never shared across customers.
Reversible
Run-level actions can be undone. A wrong task is deleted, not filed.
Questions, answered.
What is Enzo?
A model-agnostic AI agent that reads a business's data and takes action on it. It works across industries and products, inside your tenant, with a full audit trail.
Which industries and products does Enzo work in?
Enzo is vertical-agnostic. It runs in Moveezi for moving and relocation and in Taba for restaurants and hospitality, and across other Dumont products. The agent is the same; the records and the actions are configured per product.
Is Enzo tied to one assistant or provider?
No. Enzo is model-agnostic: it routes each task across models internally, and it is exposed over MCP so Claude, ChatGPT and other MCP-capable assistants can call it. It is not tied to any single assistant or chat app.
Can Enzo change my data without me seeing it?
Only at Run, and only for the capabilities you set to Run. Suggest requires a human approval before anything changes. Watch never acts. Every action is logged and reversible where the capability allows it.
Where does Enzo get its customs knowledge?
Public government sources: US CBP import regulations, US DoD DP3 personal property material and national customs authority websites, plus the documents your team uploads. Enzo keeps them in your tenant, and customs is one capability it runs deep; it also reads and acts across the rest of the operation.
Is my data used to train models?
No. Answers are grounded per tenant and your data stays in your tenant's namespace. Any licensed knowledge is never exposed to a general assistant to absorb.
It knows the work. Now let it do the work.
Start at Watch, earn Run on your own data, and put an agent on every desk.
Request access