Skip to content

Write to me.

A question, an idea, or something you just want to think out loud about. If something else would serve you better, I'll tell you.

You can also mail me directly at rasmus@tek-aerobic.dk.


The details are used to answer your request. Read about privacy.


All projects

PDF Production

From data to finished PDFs without manual layout.

AI drafts the product copy, and a person approves both the copy and the PDF.

A dataset of product data, kept up to date by a person, becomes print-ready pages. AI drafts the product texts, a person edits them where needed and approves them, and finally a person approves the PDF draft. The next edition starts again in the dataset.

Shown with example data from an invented office supplies dealer.

In and out
A dataset of product data in. A print-ready PDF out.
People
Keep the dataset up to date, edit and approve the texts, and approve the PDF draft.
AI
Drafts each product text from the dataset.
Systems
The dataset of product data. From it, the system sets up the pages as a PDF draft.

Page by page, edition after edition.

  • Every new edition means copying items and texts from a dataset into a layout, page by page.
  • Mistakes creep in along the way, and the next edition means doing it all again.

Who does what, from dataset to print.

People, AI and systems each have their own tasks. The drawing follows one new edition from the dataset to the print-ready PDF.

  1. StartProduct data kept up to dateThe owner of the product data corrects the dataset: one row per product.
  2. 1 · AIAI text draftsA draft of each product text, written from the dataset.
  3. 2 · PeopleDrafts editedThe approver can edit a draft before approving it.
  4. 3 · PeopleTexts approvedOnly approved texts go on to page layout.
  5. 4 · SystemsThe system sets up the pagesFrom the dataset, with the approved texts and the dataset's prices.
  6. 5 · PeoplePDF draft approvedA person approves the PDF draft before it is ready for print.
  7. ResultPrint-ready PDFThe edition is ready for print.The next edition starts again in the dataset.
A drawing of the steps in three lanes: people, AI and systems. Product data is in a dataset that its owner keeps up to date. From the dataset, AI drafts each product text. An approver can edit the drafts and approves the texts. The system sets up the pages from the dataset with the approved texts and the dataset's prices. A person approves the PDF draft, and the edition is ready for print. The next edition starts again in the dataset. A ring round the dot shows where a person updates the dataset. A pencil shows where a person edits an AI draft. A check mark shows where a person approves. A dot without a ring shows where AI or a system does the step. Dot without a ring: AI or a system does the step.Ring round the dot: a person updates the dataset here.Check mark: a person approves here.Pencil: a person edits an AI draft here.Dashed line: the next edition.

What each step looks like.

  1. StartPeople

    Product data kept up to date

    The workflow starts with a dataset, a list of product data that already exists. The owner of the product data keeps it up to date. In the example, each item has a row with its name, notes, size and price. Every new edition starts in the dataset.

  2. 1AI

    AI text drafts

    No product text is written from scratch. AI drafts each item's text from what is in the dataset.

  3. 2People

    Drafts edited

    The approver reads each draft, can edit it and approves the text before it goes on the pages. Drafts are not sent back, and AI does not write them again.

  4. 3People

    Texts approved

    Every text is approved by a person. No text goes to print without being read and approved.

  5. 4Systems

    The system sets up the pages

    No text is moved around a layout by hand. The system places the product names and prices from the dataset on the pages alongside the approved texts. The result is a PDF draft.

  6. 5People

    PDF draft approved

    A person looks through the PDF draft and approves it. Only then is the edition ready for print.

A print-ready PDF in which every text is approved.

The edition ends as a PDF that is ready for print. Every text and the PDF draft have been approved by a person, and the prices come from the dataset.

What changes in the work

Before

  • Items and textsCopied from a dataset into a layout, page by page.
  • Along the wayMistakes creep in.
  • Next editionEverything is done again.

After

  • Items and textsThe system sets up the pages from the dataset, and AI drafts each text.
  • Along the wayA person can edit the texts, approves them and approves the PDF draft.
  • Next editionStarts again in the dataset.

Same steps, other places.

Three invented examples of turning a dataset into a print-ready PDF. AI drafts the texts, which a person can edit before approving them. Once the pages are laid out, a person approves the PDF draft.

  • A training centre

    In the dataset
    Courses, dates and prices
    AI drafts
    A short description of each course
    A person's tasks
    Can edit each description, then approves it. Finally approves the PDF draft.
    Out comes
    A course overview for print
  • A wholesaler

    In the dataset
    Item numbers, sizes and prices
    AI drafts
    A short text for each item
    A person's tasks
    Can edit each text, then approves it. Finally approves the PDF draft.
    Out comes
    A dealer price list as a PDF
  • A conference organiser

    In the dataset
    Talks, times and rooms
    AI drafts
    A short blurb for each talk
    A person's tasks
    Can edit each blurb, then approves it. Finally approves the PDF draft.
    Out comes
    A programme for print

A list that needs to become pages?

I'll go through your message, and if the job is not a fit, I'll tell you clearly.

Describe your project

What kind of job is it?

Pick the closest one. It can change once things are clear.