marnis.ai

Mid-sized businesses

One Bot per department. Onboarded with your knowledge.

In mid-sized businesses, knowledge lives in people's heads, manuals and catalogues. A Node.Bot learns it like a new colleague: you give it the documents, it builds a second brain from them, and you practise with it before it talks to customers.

A company with two departments. Each has its own Node.Bot with its own job; both belong to one fleet and only exchange data that has been approved.

Examples

What your Node.Bot can take on.

Each example says what runs today and what comes next.

6 examples shown

Onboard it like an employee

Planned

You give the Bot access to manuals, catalogues and processes. It builds a knowledge network from them and practises with you how to deal with customers.

Example: First trial conversations with the team, then approval for customers. What it does not know for sure, it hands to a person.

Today
Knowledge networks for devices and sources, in observe mode.
Next
Onboarding with product knowledge that staff confirm. After version 1.0.

Advice on a tablet before staff take over

Planned

Customers collect ideas with the Bot in the shop or on the go. When a member of staff takes over, everything is ready.

Example: A customer is planning a kitchen, and the Bot asks about measurements and style. The salesperson sees the summary before greeting her.

Today
Nothing yet for customers.
Next
Guest access and handover to staff, after version 1.0.

Customer requests to the right place

Idea

Requests from the mailbox are evaluated and passed to the responsible department as a ticket with the key details.

Example: “Delivery request, 40 units, preferred date week 49, responsible: Sales South.”

Today
Connecting mailboxes.
Next
Sorting is in progress, passing on is an idea.

Help with coding

Today

Explain, review and write code with a code model that runs on your Node. Your source code does not leave the building.

Example: “Why does this test fail?” The Bot reads the error and suggests a fix.

Today
Questions about code in the chat.
Next
Further steps are not planned yet.

Documentation

Idea

Conversations, tickets and notes become clean minutes and instructions.

Example: “Turn today's conversation into instructions for the new machine.”

Today
Summaries in the chat.
Next
An idea: a fixed place to store and maintain documentation.

Connect departments

Planned

Several Bots, one Fleet. Sales asks the warehouse Bot, but only for what you have approved.

Example: The sales Bot asks about stock. The warehouse Bot answers with the number, not the whole list.

Today
Several Bots on one Node.
Next
Fleets with several Nodes and data exchange after approval.

Setup

What your setup could look like.

A Mac Studio per site

More memory for larger models and more conversations.

One Fleet

Several Nodes, one account, clear rights.

Onboarding

Your knowledge as a second brain.

Requirements in detail

Is this for you?

Join the waitlist. We will get in touch personally and talk about which tasks your Node should get first.

Join the waitlist