Irving-based healthcare pioneer Pieces Technologies is bringing AI to healthcare—not to replace doctors, but as a digital assistant. The company announced Tuesday that it has fully incorporated the latest generative AI services from Amazon Web Services into its healthcare solutions, supporting providers at the point of care.
The solution—called Sculpted AI—is tailor-made to meet the needs of health systems, right down to individual units, specialties, or even physicians, Pieces said.
Using the power of Amazon technologies, Sculpted AI integrates with patient records to reduce repetitive manual reviews and admin tasks. It also monitors individuals to flag risks early and reveal social determinants impacting health. For safety, it uses adversarial and collaborative AI, auto-prompt engineering, and human oversight.
“Students of healthcare continuous improvement know that the foundational principle is rapid data-driven iteration with teams on the ground,” Ruben Amarasingham, M.D., CEO of Pieces, said in a statement. “Pieces Sculpted AI brings our AI applications into that framework under extremely strict safety controls.”
“Leveraging AWS’s generative AI services allows Pieces AI roll-outs to align with a unique Kaizen methodology and PDSA principles, rather than point in time AI implementations,” Amarasingham added. “This is critical for controlled, responsible AI deployment.”
Founded in 1994, Pieces was born out of the Parkland Health & Hospital System when Amarasingham, began developing predictive models to prevent hospital readmissions.
Autonomously drafting ‘over 2 million patient summaries’
Many are worried these days about AI “hallucinating” things and producing errors or misinformation. However, Pieces says its “industry-leading” severe hallucination rate is “under 0.001%.” Earlier this year, Pieces’ system was the first to draft over 1 million inpatient clinical summaries.
So far, Pieces says it has already autonomously drafted “over 2 million patient summaries in a real-world setting.”
So how does the company use AWS to do this? It uses Amazon Bedrock, Amazon Elastic Compute Cloud (Amazon EC2), and Amazon SageMaker Canvas. The resulting advancement offers the company’s U.S. and global health system and clinical business partners “rapid, on-demand application of AI within electronic health record clinical workflows,” Pieces said.
Like having a customized ‘AI agent’
Dr. Nicholas S. Desai, MD, Chief Medical and Chief Quality Officer at Houston Methodist Sugar Land, called the speed of iteration that Pieces AI can output “stunning.”
“It’s like having an AI agent customized specifically to your organization down to the unit level,” Desai said in a statement. “They have now deployed their work at scale across several major health systems, including ours, in every age group and clinical specialty with equal acclaim from doctors, nurses, care managers, and administrators. The impact of their work with AWS is evident. Big fan of both organizations.”
Pieces has harnessed the power of AWS since its founding, driving innovative solutions and supporting major health systems nationwide.
Dan Sheeran, general manager of healthcare and life sciences at AWS, lauded what Pieces is doing with his company’s services.
“There is major opportunity and need to harness generative AI at the point of care to derive better insights and reduce administrative burden on clinicians,” Sheeran said. “In particular, being able to integrate generative AI directly into EHR clinical workflows via Amazon Bedrock and Pieces’ technology can provide meaningful value. Pieces’ continued advancements demonstrate its culture of rapid innovation and responsible approach to AI for the benefit of both patients and clinicians.”
Leveraging AI is just one of the ways Pieces has impacted U.S. healthcare recently. In 2022, the company partnered with Philadelphia-based nonprofit research firm NRG Oncology to accelerate patient identification in a cancer study that could impact millions.
Get on the list.
Dallas Innovates, every day.
Sign up to keep your eye on what’s new and next in Dallas-Fort Worth, every day.