Frequently Asked Questions

Who is behind Prodigy?

Prodigy was developed by Ines Montani and Matthew Honnibal of Explosion AI. We're also the makers of spaCy, the leading open-source library for Natural Language Processing in Python, and several other open-source tools.

How is Prodigy different from other annotation tools?

Prodigy is designed to fit neatly into a modern data science workflow. It emphasizes productivity and extensibility. To make annotation more efficient, Prodigy supports continuous active learning. This means that the model learns as you annotate and only asks you question it doesn't already know the answer to. The annotation interface is a lightweight web application that follows modern UX principles: focus on one decision at a time, reduce friction and automate as much as possible.

Prodigy is a tool that you can download, install and run yourself – not a service. This has several benefits: you can interact with it programmatically, import and extend it, and you don't need to upload anything into the cloud.

How much does Prodigy cost? Do I have to buy a subscription?

Prodigy is available as a downloadable tool for a one-time fee with 12 months of free upgrades. You can choose between two different license types: a personal license or a number of transferrable company seats. See this page for more details and pricing. For the license agreement, see our terms and conditions.

What does the added VAT mean?

Our company is based in Berlin, Germany, so if you're ordering from within the EU, you will be charged an additional VAT. If you're a VAT registered business in the EU and not based in Germany, the reverse charge mechanism applies, meaning no VAT will be added on checkout. Businesses and individuals outside the EU won't be charged an additional tax.

Do you offer any plans specifically for universities?

Yes, universities and degree-granting research institutions can purchase an institutional license for Prodigy that includes 12 months of unlimited usage for all students and staff. Researchers can also apply for a free, 3-month interim license in order to keep working while the institution is getting set up with a site-wide license.

Why is the academic license limited to degree-granting institutions?

Today's academic community includes an increasing number of people employed by large companies. This makes defining a "researcher" difficult. We've limited the academic license to degree-granting institutions such as universities as a simple way to define a "bright line". Universities also have some special concerns. For instance, without a site-wide license, using Prodigy in an undergraduate course would mean issuing and revoking a large number of temporary licenses.

How can I get help and support with using Prodigy?

Prodigy comes with extensive documentation, including API specs of the individual components, examples of custom recipes and data formats, as well as various tips and tricks for achieving the best results. We're continuously working on providing more, freely available resources and educational materials like video tutorials and usage workflows. We also answer user questions on the Prodigy Support Forum.

Which platforms and operating systems does Prodigy support?

Prodigy is compatible with Python 3.5+ on macOS / OSX, Linux and Windows. The web application can be accessed from any web browser, on both deskop and mobile or tablet devices.

Can I use Prodigy with other machine learning libraries like TensorFlow?

Yes! Prodigy follows a component-based architecture and is built around recipes – simple Python scripts that control the processing logic and text source, and specify how the tool interacts with your model. The built-in recipes are powered by spaCy, Thinc and LightNet – but you can easily write your own, using any framework or library of your choice. For example, an image classification workflow with Keras, or a language generation system with PyTorch. We're also working on extending Prodigy's built-in support for other machine learning libraries.

Can I extend Prodigy with more interfaces and use it to manage annotation projects with multiple annotators?

Prodigy is mostly designed as a developer tool that runs on one machine. However, you're free to extend the tool with any capabilities you need, and choose how to serve it. The web application is fully fully themable and lets you use custom HTML templates to visualise your examples in any custom way you want.

We're also working on an extension library, the Prodigy Annotation Manager, which will integrate with Prodigy and will let you set up complex annotation projects, manage multiple annotators, enforce quality control and keep track of the progress via an admin console. You can sign up to our mailing list to be notified about the private beta.

Why aren't you offering Prodigy as a service? Are there plans for a SaaS version?

Software as a Service is a very popular business model, because it gives repeat revenue and a simple story for raising capital. However, not all software products are most useful as a service. Prodigy is designed specifically for developers and data scientists. We believe that it's most powerful if you can run it yourself, configure it with code, extend it however you like and keep the data on your servers. That said, we do see potential for a slightly different SaaS version of Prodigy, particularly for larger scale annotation projects.

scikit