Do you remember the scene from “Her” where Samantha, the AI, goes through Theodore’s letters and classifies them into Amazing and not-so-amazing?
While the AI solutions provided by natif.ai don’t mend broken hearts as Samantha does, they can definitely classify your documents for you!
At natif.ai, we understand that as a prosperous business, you are most likely facing a flood of documents from different sources without knowing what they contain and who should handle them. It probably requires intensive and tedious manual work for you to go through each document and decide whether to toss it away, pass it on to a co-worker, or sort it into a specific folder.
If this applies to you, then you can stop worrying as natif.ai has launched a service where you can train your own AI model on your own documents to sort them into your own chosen labels! Couldn’t get more convenient, ha?
Let’s Start
We start in the
Workflow Overview of our platform and choose “
Train Your Own Model Now”.
Select Your Workflow
Here you can find all our Custom AI Workflows. For our Classification Workflow we select “Create Custom Classification“.
Describe Your Custom Classification Workflow
We start with describing our workflow by giving it a name and short description. You can also upload an image. This will help you to distinguish this workflow from the others.
Define The Workflow Classes
Now you can tell the model your desired classes. It’s as easy as just naming your classes! If you provide descriptive names for your classes, our generic classification model can even work out of the box without training. Naturally, however, you will get better classification results once trained on your data.
In our case we want to sort invoices and receipts.
We also add the class “others” for any other document types that might appear.
Specify Your Documents
Now we have to give the AI some information about our documents so it knows which tasks need to be done. This also improves the accuracy of your workflow.
For a Custom Classification Workflow the AI needs to know:
– Are the documents always perfectly cropped or should they be cropped in the workflow?
– Is the content on the documents in the Latin or Japanese alphabet?
– Is the text printed or handwritten? Or can it be both?
– What pages are needed to categorize the documents?
– Should the documents be classified based on text or visual features?
Your Workflow Is Created
Your model is ready but is initially just a simple generic one. It still needs your guidance to excel!
Right now, you can test the model on your own data, however, the results might not be as satisfactory because the model will behave based on its generic knowledge. If your data poses unique challenges, you will need to give it some training first. For this we select “Upload Training Data“.
Upload Your Training Data
Now, it’s the long-awaited moment for your model. It is going to learn from your own data so its results can be tailored to your documents specifically.
If you have already sorted some files into classes (or can get class labels from your existing databases), then just give these sorted files to the model to directly learn from.
Here, you can first select the name of the class, then upload the files that belong to such a class.
Please upload a minimum of 50 documents per class. It’s very important to select documents that are very similar to the documents that the model will process later.
This will help the AI get a full understanding of your documents and provide high accuracy processing.
If you upload the documents already sorted into classes, you don’t need to annotate them. The AI will already know which class the document belongs to.
If you upload your documents unsorted, you have to annotate them.
Annotate Your Training Documents
For annotation your unsorted documents, an uploaded document will be shown with the probabilities for each class as computed by the generic model. If the generic model is wrong, just select the class that you see fit, and the model will adjust when you start training.
Now repeat this step for each of your uploaded documents.
Start The AI Training
Once you are done with annotating the documents, you can start the training.
This means the AI now learns how to process your documents.
You will receive an email once the training is completed – which is normally within the next 24 hours!
Integrate Your API
However, your workflow API is already ready and can be integrated! You can find all information such as code snippets and json response examples in the workflow documentation.
That’s It!
Your API will automatically be adjusted when the training is completed! The training metrics will provide you with more information about the accuracy of your AI Workflow.
If what you want to do next is more complex than just classifying documents, then, check out our
Workflow Overview and see if it already contains what you want to do! Even if it doesn’t, just
contact us and let us how we can support you.
