Document processing for developers

The best way to extract structured data from unstructured documents.
Harness AI to process documents at scale.

Backed by
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Companies of all sizes trust automaited's document extraction platform.

Integrate today

Just define your target data structure as a good old Pydantic model. Descriptions of fields are used to guide the LLM during extraction. Nesting and custom validation methods are also supported. Try 1000 free extractions with the snippet blow. Learn more.
pip install automaited
from datetime import date
from pydantic import Field, BaseModel
from automaited import DocExtClient

class Article(BaseModel):
    article_number: str | None = Field(None, description="Typically alphabetical or alphanumerical.")
    description: str | None = Field(None, description="Description of the item.")
    quantity: float | None = Field(None, description="Number of pieces.")

class PurchaseOrder(BaseModel):
    customer_name: str | None = Field(None, description="Examples: Kaladent Inc., Henkel GmbH")
    order_number: str | None = Field(None, description="The purchase order number.")
    order_date: date | None = Field(None, description="The purchase order date.")
    items: list[Article] = Field(default_factory=list, description="List of all ordered articles.")

client = DocExtClient(API_KEY="TEST_BETA:you@company.com") # Replace with your email, prod API keys coming soon
result: PurchaseOrder = client.extract_model(PurchaseOrder, "./purchase_order.pdf") # automaited.dev/samples
print(result)

Automatic UI for your data structure

We automatically generate a form using your target Pydantic model. This form can be used to adjust extraction results before further processing and to enforce compliance. Adjustments in the data are stored as test data to help you improve your extraction model. Adjusting the UI and adding validation can be done through your Pydantic model. Learn more.

Ingest documents via API and Email

Your clients can directly trigger your extraction model via an email endpoint or you can send the documents via API. You can also respond with an email containing a link to the extraction. Learn more.

Deploy easily

Click a button to deploy your extraction model or automatically deploy on merge into a predefined branch. If you want to run your extraction model on your own server you can do this too. Learn more.

Integrate easily though Webhooks

We offer a few standard integrations to allow your clients to transfer data to their target system. But you can also build your own, just use our webhooks to get informed when your customer clicks the "transfer"button in our UI or when an extraction is tirggered in general. Learn more.

Beta launch

Join mailing list

Try today, get updates quickly. We don't spam.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Update cookies preferences