Demos
Enhancements to the AI and NLP technology also resulted in tripling the company’s customer base over the last year
John Snow Labs, the AI and NLP for healthcare company and developer of the Spark NLP library, today announced that it has reached the five million download mark, doubling the growth of its flagship natural language processing library since January. This growth can be attributed to a long series of updates and enhancements, with new software releases made every two weeks for the past 40 months consecutively. Recent upgrades include significant gains in speed and capabilities, including richer medical named entity recognition and multi-modal learning to classify documents using both text and visual layout.
Over the past 12 months, the company has tripled the number of paid customers who license its commercial editions: Spark Healthcare NLP for analyzing clinical and biomedical text, and Spark OCR for analyzing visual documents. New customers include four of the world’s largest pharmaceutical companies and two of the nation’s largest healthcare companies. Used by 33% of all enterprise AI teams and 54% of healthcare AI teams, Spark NLP has secured the spot as the most widely used NLP library in the enterprise.
Spark NLP for Healthcare has established new state-of-the-art accuracy in peer-reviewed papers on public academic benchmarks over the last year. This includes providing the most accurate models ever for medical named entity recognition, relation extraction, assertion status detection, and adverse drug event detection. In contrast to academic research results, John Snow Labs’ implementation is production-grade, natively scalable, optimized for the latest CPU and GPU chips, easily trainable, and hardened for the security and compliance requirements of healthcare organizations.
With a set of more than 1,400 pre-trained NLP models supporting 200+ human languages, Spark NLP is reusable with a few lines of code, further democratizing the use of AI and NLP technology across the globe. The latest 3.0 release is helping John Snow Labs continue to deliver on this promise and fuel its rapid growth. Highlights include:
- Massive Speedups – Spark NLP 3 is now 8x faster for calculating BERT-Large and 3x faster for calculating named entity recognition. This follows months of profiling and optimization to get the most from modern hardware, whether the technology is running on CPU or GPU, locally, or on a cluster.
- New Medical Named Entity Recognition – Spark NLP for Healthcare 3 includes pre-trained models that automatically identify over 80 entity types. This expands support for out-of-the-box identification of social determinants of health, imaging results, demographics, vital signs, and clinical events.
- Multimodal Learning to Classify Text and Visual Layout – Spark OCR 3 applies trainable, scalable, and production-grade implementations of LayoutLM. This enables highly accurate information extraction from visual documents such as lab reports, invoices, clinical trial records, next-generation sequencing reports, and other images that rely on visual cues in addition to text.
“Trends including the urgent industry need for more personalized medicine and accelerating clinical trials, paired with clinicians’ continued reliance on free text and new AI breakthroughs is making NLP a foundational healthcare technology,” said David Talby, CTO, John Snow Labs. “We are proud to help our customers accelerate the progress of healthcare by translating the most recent AI research advances into production-grade software that continues to exceed industry standards for accuracy, scalability, and privacy.”
As part of John Snow Labs’ commitment to delivering the best NLP solution to the global AI community and healthcare industry, the company fosters an active community of knowledge sharing. Last month, in partnership with Gradient Flow, the company issued the AI in Healthcare Survey Report, exploring the state of AI in healthcare. The research was released in conjunction with the inaugural Healthcare NLP Summit, a free, virtual event that brought together 4,000 of the brightest minds in applied AI for healthcare. Coming off the heels of its success, John Snow Labs will host its second annual NLP Summit from October 5-8. Free registration is now open.
