Tuesday 12pm, 14 February 2017
Scaling Visual Investigations with Graphs, GPUs, Design, & Automation
CEO - Graphistry
From hunting hackers to stopping outages to fighting fraud, investigation tools must scale to meet the demands of the data era. This talk presents the design and implementation of Graphistry, a visual investigation platform that is now piloting at some of the world's largest organizations.
Interesting in its own right, we will first overview ways that human-in-the-loop investigations are hard for even the best-funded teams in the world. From there, we will present four ways Graphistry makes investigations faster and more reliable in high-scale settings:
- Scaling interactive visual analytics by connecting browsers to GPU clusters
- Enabling collaborative workflow automation through visual playbooks
- Presenting high-dimensional data in a rich, coordinated visual graph analysis environment
- Supporting high-density visual designs through algorithmic techniques
Put together, Graphistry enables faster, easier, and more reliable investigations over 10-1000X more data.
Leo Meyerovich co-founded Graphistry, Inc. in 2014 to scale and streamline visual analytics. Part of Graphistry's approach builds upon the founding team's work at UC Berkeley on the first parallel web browser, GPU-accelerated visual analytics, and program synthesis. Leo's most referenced research is in language-based security for the web and access control. Previously, he worked on the first parallel web browser (PLDI SRC, Qualcomm Innovation Fellow), sociological foundations of programming languages (OOPSLA best paper, SIGPLAN best-of-year highlight), and while at Brown University, the first functional reactive web language Flapjax (OOPSLA best paper, NSF GRFP).