Celebrating Our Free & Open-Source Contributions in 2022

As we enter 2023, John Snow Labs' focus is on Responsible AI, empowering domain experts, and giving you control over large language models.

Thank you for supporting John Snow Labs in 2022. With the help of a global AI community, we’ve been able to grow our free & open-source initiatives and have several milestones to celebrate together this holiday season:

• Spark NLP crossed 5 years of releasing software every two weeks. It has over 45M downloads of the open-source library across 12 supported platforms.

• The Annotation Lab is now actively used by more than 100 organizations worldwide to label, train, and test NLP models without coding.

• Spark NLP for Academia provided almost 300 free academic licenses to accelerate research into Healthcare NLP to the universities of Harvard, Cambridge, Stanford, Oxford, MIT, Duke, Toronto, Georgetown, Madrid, Chicago, Milan, Athens, and many others.

• The NLP Summit once again drew more than 10,000 people this year, as the largest gathering of applied NLP professionals worldwide. Other educational initiatives include the Healthcare NLP Summit, monthly webinars, talks at over 20 conferences, and over 500 free student passes to Spark NLP training & certification.

• Peer-Reviewed Papers. The 9 papers we co-published this year established new state-of-the-art accuracy in clinical relation extraction, adverse event detection, and biomedical named entity recognition. The team also contributed to medical research papers on COVID-19 outcome prediction and risk factor detection.

In our commercial work, we’ve significantly expanded the Healthcare NLP library which now has 800+ state-of-the-art pre-trained models including industry-first models specific to oncology, radiology, public health, and social determinants of health. We’ve also launched Finance NLP (100+ models & pipelines), Legal NLP (500+), and Visual NLP (expanding Spark OCR with Document AI capabilities). Ease of use is greatly improved with one-click deployment on AWS, Azure, and Databricks, and a new Python library with simplified installation and access to thousands of models with one line of code.

 We’ve grown our revenue and customer base – which enables us to keep you at the state of art of NLP & AI for years to come, whether you use our paid products or free software. This primarily means providing you with production-grade versions of the most accurate models ever, tuned for your needs. But going into 2023, it also means other things. We need to make Responsible AI a reality and raise the bar of what a production-ready model means. We need to enable doctors & lawyers instead of data scientists to train AI models. And we need to make the new class of large language models work for you under your control. As always, we’d appreciate your help by telling us what to build for you, trying out the new software, or contributing to the open-source codebase.

 With Thanks, Happy Holidays, and a Happy New Year!