Marius Eriksen Interview
summary
GRAIL detects cancers early and pipelines generate up to a terabyte per sample, forcing investment in storage, network and cost-aware compute. Analysis runs on cloud providers with heavy use of the Amazon spot market and clever job scheduling to cut costs. Statistics and modeling use R, while a custom workflow language called Reflow (now open source) expresses dataflow-aware pipelines so the runtime can optimize I/O and parallel execution.
AI-written summary · about
related talks
50mCancer Genomics - a biologist and a developer
07mCloud architecture for population-scale analyses of genomes and AI.
10mGenesis Final Round - Drylab AI
27mGen-AI at Scale: Simplifying Orchestration of Healthcare Applications Across Multi-Cluster Environme
11mEnsuring Reliability in Healthcare Data Engineering
46mUsing Open Source Tech to Swap Out Components of Your Data Pipeline
29mBuilding Self-Serve, Multimodal Medical Image Pipelines with Ray
01h 03mMerck drives innovation with a “real-world data exchange”