Custom Extractors
Custom models can be created within a matter of weeks to support document types that are not currently supported by Affinda's base extractors.
Customers can choose any document type they want to automate, and the Affinda team will work to develop a custom base extractor for it. See below a non-exhaustive list of example document types that could be automated:
Building your own custom model with Affinda
To ensure that customers are able to implement an effective automation solution, Affinda offer 4 key services to help build and operate your custom model:
- Scope
Align on end-to-end solution design, including feature requirements such as classifications, extraction models, validation rules, required APIs, and whether self-learning capabilities are required.
- Build
To build a model, we require a sample of at least 50 documents per document type for training purposes. Depending on the document complexity, up to 250 documents may be required to develop a performing model.
- Pre-launch
Affinda assists in planning, testing, and training for a seamless transition to the live model.
- Operate
We offer Hypercare support to swiftly resolve operational and production issues in the period following model ‘go-live’.
Post-deployment, Affinda will provide ongoing support.
Setting up your custom models
To discuss further whether tailored models are appropriate for your use case and to understand further how to best configure your account to take advantage of them, contact us to set up a discussion so that we can understand your use case better and provide some recommendations.
Updated 12 months ago