Before
Documents arrive by email and get renamed, sorted and filed into the right folders by a person.
After
They land in the right place, named the right way, the moment they arrive.
Documents arrive by email and get renamed, sorted and filed into the right folders by a person. They land in the right place, named the right way, the moment they arrive.
Documents arrive in a shared inbox all day. Someone opens each one, works out what it is and who it belongs to, renames it to match the convention, and drags it into the right folder. In the companies we meet this is a constant background task that nobody owns and everybody does.
Files get misnamed. Folders drift. Six months later nobody can find the document that was definitely filed.
Call a fully loaded employee $45,000 a year, as an illustration. Filing typically takes 5 to 10 hours a week across a team. Ten hours is a quarter of a person, spent dragging files.
Typical figures, using $45,000 as an illustrative fully loaded cost. Your numbers go in the calculator below. If your version is a couple of hours a week, that still counts.
We build an automation that watches the inbox. When a document arrives, it reads the sender, the subject, the attachment name and, if needed, the contents. It renames the file to your convention and files it in the right folder. The person who sent it can get a confirmation.
Where the naming rules are clean, this is plain automation. Where they are not, and a person would have to read the document to know what it is, we use AI to classify it. We tell you which documents go down which path, and anything the classifier is unsure about goes to a person.
Plain automation handles most repetitive work. We use newer tools only where they clearly earn their place, and we tell you when we do. Here, classification is the one place they might.
What the team gets back is the inbox. Documents are where they should be before anyone has to look for them.
Is this AI?
Mostly plain automation. If the sender, the subject or the file tells you where it goes, that is a rule. If a person has to open the document and read it to know, we use AI for that one step, and we tell you.
Automation vs. AI: which one your task needsBefore
Documents arrive by email and get renamed, sorted and filed into the right folders by a person.
After
They land in the right place, named the right way, the moment they arrive.
Real work from real clients, with the names left out. What matters is the shape of the problem, and how often we see it.
This kind of project
Depends on the naming rules
When the sender, the subject or the file name tells you where a document goes, this is a small project. When someone has to open and read the document to know, we add a classification step, and that takes more building and more checking. We look at a sample of real documents before we recommend which.
Set to the typical figures for this task. Change them to match your team, however small the number.
Hours a week, times the people doing it, at whatever a person costs you. $45,000 a year is an illustration; set it to your own figure.
People doing this work
1
Hours each, per week
Half an hour counts. 40 is a full week.
10
Fully loaded cost per person, per year
Salary, benefits, taxes, overhead. $45,000 is an illustration; set it to yours.
$45,000
Time spent on this task
10 hrs / week
About 520 hours a year of a person's time.
These figures show the value of time spent on the task, not a project quote or guaranteed cash savings. We will help estimate what can realistically be automated.
Every morning, every Friday, or every time an order comes in. Small tasks and larger workflows are both welcome. The evaluation is free, and if we take it on, first deliverables are running in your operation within 30 days of kickoff.
Free evaluation. Fixed quote before work begins.
Automation, AI where needed
Stop copying email contents into a spreadsheet by hand
The row appears on its own. Nobody retypes what was already written.
Plain automation
Automate preparing files for someone else to review
The package builds itself to the same standard, and the reviewer is told when it is ready.
Everything we automate
Fourteen tasks, each with what it takes today, what we build instead, and what shapes the scope.