How data science shortened the COVID-19 pandemic by 2 months
summary
Claims that data-driven geospatial and predictive models cut U.S. vaccine trial time by 6–8 weeks, supported by concrete numbers about enrollment speed and exposure rates. Workflows combine county-level forecasts, rapid targeted recruitment into emerging hot zones, and monitoring of demographic representativity. Implementation details include productionized pipelines, a serverless fleet for parallel LLM tasks, and tooling tied to Snowflake Whisper OpenAI and Bedrock for simulation, transcription, and rapid analysis.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:07County-level COVID spread
- 3:56Vaccine trials shortened
- 5:19Flaws in low-code AI
- 7:07Collaborative SEIR modeling
- 9:53Data science challenges
- 11:42Zerve’s platform
- 13:07Dependency management
- 14:32LLM call center data
- 17:42AI call transcript analysis
- 21:38Compliance violations detection
- 23:50Agent performance AI
- 25:15APIs without cloud
- 27:01Zerve’s free tier
- 28:05Live demo & Q&A
related talks
22mCOVID-19: Data Science & Expertise
23mApplying AI and ML in the Current Pandemic World
34mFighting COVID with Python
27mBeyond a Machine Learning Model
41mScaling sentiment: Using digital surveys to geotarget vaccine distribution
28mThe COVID Scenario Pipeline: High Stakes Data Science
37mAI/ML Strategies & Techniques for a Post Pandemic World
31mAn incomplete list of implementing data science effectively