<!-- Markdown-versie van https://outdeliver.ai/en, gemaakt door tools/bouw.mjs -->

AI in your own Microsoft environment

# An ordinary Tuesday. This time with AI.

Follow a day in an organisation that started with AI a year ago. Seven
kinds of work that come up almost everywhere. All of it is possible
today, and at every step people stay in charge.

[Book a call](https://outdeliver.ai/en/book-a-call.html)
[Get started](https://outdeliver.ai/en/get-started.html)

A Tuesday, a year from now

## Follow the day

For each moment you see who decides and what you measure. We measure
the work, not individual employees.

08:00 In the sales office

Incoming email

### The inbox is already sorted

Orders, invoices, questions and complaints arrive in one inbox or
ticket system. AI reads them, enters each order in your system as a
draft and sends the rest to whoever needs it. Complaints, for
example, get priority.

Who decides

Inside sales checks the draft orders and passes them on. Anything the AI is unsure of is set aside with an explanation.

What you measure

How long a message waits and how many orders go through without rework.

Requires: access to the inbox and your ERP

09:15 In customer service

Customer service

### A customer question, with the full picture

A customer asks where her delivery is. Everything is on one screen:
what was agreed, the order status and earlier contact. A draft reply
is ready, with its sources alongside. No chatbot talks to the
customer directly.

Who decides

An employee sends every reply. Complaints and exceptions get an extra check.

What you measure

The handling time and how often a customer comes back with the same question.

Requires: your CRM, the order status and the mailbox

10:30 On the phone

Phone calls

### After every call the summary is ready

A colleague talks to a customer about a delivery. By the time they
hang up, what was agreed and the follow-up tasks are already in the
customer file. Nobody has to write up notes any more. The customer is
told at the start that the call is being transcribed. You agree in
advance who can read the transcript and how long it is kept.

Who decides

The colleague reviews the summary and passes on the tasks.

What you measure

The time still spent after each call and how often an agreed action is not followed up.

Requires: your phone system and the customer file

11:30 In sales

Sales and quotes

### The quote is ready

A request comes in through the website. AI finds the right products
or services, prices them on the terms agreed with this customer and
prepares a quote in your own template.

Who decides

An employee reviews the quote, approves it and sends it.

What you measure

The time until the quote is sent and the number of corrections.

Requires: your pricing agreements and the quote template

13:30 In marketing

Marketing

### A finished project becomes a story

When a project is finished, AI prepares drafts for a post, the
newsletter and the website. They come from the project report, in
your own tone and house style.

Who decides

Marketing chooses what is published and edits it. A customer name or photo only goes out with the customer's approval.

What you measure

The time per piece and how much is actually published.

Requires: project reports, earlier posts and your house style

15:00 In accounts

Administration and finance

### Every invoice already matched to its order

Every purchase invoice from today has been checked against the order
and the agreed terms. Whatever matches is ready to post. Anything
that differs or arrived twice is set aside with an explanation.

Who decides

Accounts posts and approves. Payments are always made by a person.

What you measure

How long an invoice waits and how many errors only come to light afterwards.

Requires: the orders and goods receipts from your ERP or accounting system

16:30 At his desk

HR and knowledge

### Colleagues find the answer themselves

A colleague wants to know how much leave he has left and what the
rules are on working from home. He puts the question to the staff
handbook and gets an answer based on your own policies, with the
source alongside.

Who decides

Leave is approved by the manager. Decisions about people, such as recruitment or appraisal, are always made by a person.

What you measure

How often HR gets the same question.

Requires: your staff handbook and the leave balance

22:00 While everyone is asleep

Operations and privacy

### Sensitive data stays in your own environment

Overnight, an AI agent reads the email that came in after five and
drafts the orders and replies, ready for an employee at eight. Every
step is logged. What each AI may see is agreed in advance: the more
sensitive the data, the closer to home the model runs. Once a fixed
daily cost limit is reached, the agent stops.

Who decides

For each application you decide what the AI may see, together with your IT partner. At night the AI only sees what that one task needs.

What you measure

What it cost and where a person stepped in.

In the Netherlands, 49 per cent of employees use AI for work at least occasionally through a personal account or app (Alert Online 2026, 378 employees). Your own environment gives them a safe place to do so. For anything outside it, you set clear rules.

Requires: your own Azure environment, in a European region

- **Public** Website, brochures and product information. A European cloud service is enough.

- **Internal** Quotes, contracts and schedules. Not used to train a model. In your own Azure environment, in a European region.

- **Personal** Personnel files and identifiable customer data. Only in an isolated part of your own Azure environment, or on your own hardware if you prefer.

**A person decides.** The AI does the groundwork. Nothing goes out without an employee's approval.

Worked example

## How much time is freed up?

Take work that comes up every week, such as retyping and checking.
Move the sliders and see how much time it costs.

1,150 hours freed up each year for other work

That is **€46,000** in wage costs a year.

This work now costs **2,300** hours and **€92,000** a year.

10 employees × 5 hours a week × 46 working weeks × €40 an hour, then times 50%

A worked example, not a promise. Set against it are the build and the monthly costs, which we quote in advance as a fixed price. Together we work out what you can really gain in your own work.

[Work out your own figures](https://outdeliver.ai/en/book-a-call.html)

Three ways to start

## What will your Tuesday look like a year from now?

Start small, with one process that comes up every week. Choose how:
from half an hour behind the scenes to a session at your offices.

### Look behind the scenes

Free

Half an hour, by invitation

See everything we build running in our sandbox, the Proeftuin. No sales pitch, unless you ask for one.

[Request an invitation](https://outdeliver.ai/en/microsoft.html#toegang)

02 Recommended

### Introductory call

Free

Half an hour on Teams

You tell us what you already do with AI. We tell you what fits, or whether it is still too early.

Call with Marcel Logtenberg

[Book a call](https://outdeliver.ai/en/book-a-call.html)

### Get started

From €495

At your offices, excluding VAT

An AI quick scan, an AI assessment or an AI assessment plus inspiration session, at a fixed price.

[Get started](https://outdeliver.ai/en/get-started.html)
