Skip to content

Multimodal AI

Room photos and documents as model inputs, checked before anyone trusts them.

visionimage understandingdocument extractionvalidation

Lessons

  1. 31.1Vision models: a hotel agent that reads room photosAn AI model can look at a room photo and tell you if the bed is made. Here's what it sees well, where it fails, and when a human decides. 30 min
  2. 31.2Document extraction: invoices, passports and contracts to structured dataTurning invoices, passports and contracts into clean data: the AI reads, your code checks, and people only look at what doesn't add up. 30 min

On the job: try it on your own project

Find one manual task at work that involves images, PDFs or scanned documents, such as checking invoices or photos. Ask Claude Code to prototype extraction on ten real samples on a branch, returning structured fields. Check every field by hand. Write a short feasibility note: accuracy, failure cases, cost per document, and where a human must still review.

Can you do this now?

0/4. Anything you can't tick yet: reopen that lesson's practice ladder.