BARCELONA, Spain--(BUSINESS WIRE)--DistillerSR Inc., the market leader in AI-enabled literature review software and creator of DistillerSR®, today announced the availability of Smart Evidence Extraction (SEE)™, the industry’s first purpose-built GenAI solution for literature reviews. Depending on the subject matter, researchers generally take 2-3 hours to critically read and appraise a scientific journal’s findings, which, over the course of hundreds and thousands of references analyzed during a literature review, can lead to human fatigue and errors. With SEE, it now takes just a few seconds to extract and link the supporting evidence from scientific papers.
SEE is a composite AI technology, blending traditional AI and GenAI models. This approach addresses the information overload and context-noise issues faced by traditional GenAI models, providing faster, higher quality answer suggestions. Integrated into DistillerSR’s auditable workflow to ensure complete regulatory compliance and efficient human-in-loop oversight, SEE automatically extracts content to answer both closed and open-ended questions, including numerical content, sentiment analysis, risk of bias information and study summaries. SEE’s composite AI approach grounds the creativity of current GenAI models restricting response content to only the paper being reviewed.
“We’ve incorporated AI into the DistillerSR platform since 2016,” said Peter O’Blenis, CEO, DistillerSR Inc. “We view AI as part of our technology stack, integrated seamlessly within DistillerSR’s powerful workflow and evidence management platform, to dramatically accelerate the literature review process while still addressing the critical regulatory requirements of our customers. SEE is the next logical step in the evolution of AI in literature reviews.”
SEE incorporates comprehensive controls to manage copyright adherence, while ensuring that all data remains securely within the DistillerSR platform and is not shared with third-party GenAI vendors. In September of this year, DistillerSR adopted the NIST AI Risk Management Framework (RMF) to govern its AI models and third-party vendors. The framework allows customers and prospects to evaluate how the company governs the engineering, privacy, and ethical dimensions of newly developed AI features and their application to enhance health research.
Literature reviews are the cornerstone of evidence-based research; however, their production has traditionally been highly manual, time consuming, and error prone. In recent years, technology companies have attempted to apply large language models (LLMs) to extract and summarize scientific literature and clinical data. Most LLMs, however, generate significant inaccuracies, often called "hallucinations”, which can lead to serious data errors in regulatory submissions, scientific communications, and guideline development.
SEE provides researchers with a purpose-built GenAI solution that is integrated into an auditable and efficient human-in-the-loop literature review workflow. SEE’s proprietary technology grounds the creative nature of GenAI, making it more context aware and focused only on extracting data specific to the literature being reviewed.
From November 18-20th, DistillerSR will demonstrate SEE to ISPOR Europe conference attendees at booth 512. ISPOR is the Professional Society for Health Economics and Outcomes Research. The non-profit organization is the world’s largest professional society for health economics and outcomes research professionals.
Today, 80% of the world’s largest pharmaceutical and medical device companies trust DistillerSR to securely produce transparent, audit-ready, and regulatory compliant literature reviews. The DistillerSR platform and modular ecosystem enable our customers to securely automate the management and analysis of evidence-based research — faster, more accurately, more transparently, and at scale.
About DistillerSR Inc.
Visit www.distillersr.com.
® DistillerSR Inc. is a registered trademark of DistillerSR Inc.
™ Smart Evidence Extraction is a trademark of DistillerSR Inc.
