Accelerating COVID-19 insights for clinical use with intelligent search
summary
PanSurg builds a living systematic review called REDASA that pipes targeted web crawls into S3 and Amorphic, uses MirrorWeb for capture, SageMaker for labeling, and Amazon Kendra for natural‑language search and extractive QA. The pipeline labels with a human-in-the-loop workflow, curates roughly 1,500 documents in two weeks across ~15 COVID queries, produces autonomous summaries, and opens the curated dataset for veracity checking and downstream ML — a working prototype that materially speeds evidence synthesis.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:45Critical information is getting harder to find
- 1:50Creating great search is costly and complex
- 3:32The AWS ML stack
- 4:15Amazon Kendra
- 9:31Advantages of intelligent search
- 11:43Employees and customers want answers, not choices
- 12:41PanSurg - Research innovation during a pandemic
- 15:25Real-time Data Synthesis and Analysis
- 17:00The search challenge
- 18:06Implementing intelligent search
- 21:10Measuring the impact
- 24:28Learn ML with AWS Training and Certification
related talks
26mAccelerating COVID-19 insights for clinical use with intelligent search
43mHarnessing AI to strengthen trustworthy information
42mBuilding Production RAG Systems for Health Care Domains
45mDigital Disruption in health & medical research
38mAI-Ready Data Foundation Using Amazon OpenSearch
50mInnovation and data analytics: Solving real-world challenges
57mHow Machine Learning and AI Can Support the Fight Against COVID-19
59mArtificial intelligence in Healthcare