The AI Institute at Schmidt Sciences is a grantmaking and research group that views AI as a transformative force for scientific discovery and societal progress. Over the next decade, we aim to support key researchers who are working to make AI systems competent, trustworthy, reliable, and able to effectively partner with human scientists on the next generation of discovery.  We will also make distinctive investments in beneficial AI areas where philanthropy has a unique advantage.  By supporting enabling infrastructure, foundational research, and targeted programs in science disciplines, the AI and Advanced Computing Institute will create the conditions for AI-enabled discovery to achieve its promise.

 

The AI Institute currently has three focus areas:

 

AI for Science – using AI to improve how scientists generate hypotheses, conduct experiments, analyze data, and produce new knowledge – and do this in a way that specifically accelerates the discovery process.  Exploratory programs include post-transistor hardware for AI and AI for scientific simulation.

Science of AI – understanding and controlling AI systems to mitigate and manage potential risks from advanced AI.  Improve AI reliability and performance in areas of limited commercial interest.  Existing programs include AI2050 and Science of Trustworthy AI.  Exploratory programs include AI interpretability and evolution of multi-agent communication.

Beneficial AI – providing scientific foundations and datasets for understanding the larger human impacts of AI.  This includes selected high-impact grantmaking, such as our programs to use AI to accelerate humanities research (Humanities and AI Virtual Institute), and to quantify the impact of AI on the labor market (AI@Work).

 

These programs support and draw on related Schmidt Sciences engagements, such as the Virtual Institute for Scientific Software (VISS), which seeks to accelerate the pace of scientific discovery through the support and development of high-quality, community-oriented scientific software.  These philanthropic efforts support experienced engineers who are tasked with building open-source infrastructure and applications for multi-disciplinary, high-performance, at-scale scientific discovery.

 

Reporting to the Compute Program Scientist in the AI Institute, the Infrastructure Engineer will provide technical leadership and apply computational infrastructure management expertise across AI Institute and Schmidt Sciences multidisciplinary efforts.  The initial set of projects will focus on cutting-edge developments in AI, and a successful candidate should have a work portfolio that reflects specific contributions to deployment and management of appropriately-sized heterogeneous compute clusters in a research or commercial environments.  Success for this role is defined by the adherence to the industry’s best DevSecOps practices at scale, and the ability to quickly address computing needs from multiple research teams. This role requires up to 50% domestic travel.

 

Example activities for this role:

●      Implementing multiple Authorization and Authentication schemas to accommodate a diverse set of cluster users and applications.

●      Managing a mixed-type network storage system with different access models and hardware performance characteristics.

●      Designing, implementing, deploying, and maintaining a Continuing Integration Continuous Delivery (CICD) system to address the needs of geographically distributed developers and applications.

●      Integrating multiple types of workflow orchestration systems in combination with software dependency package and module management frameworks.

 

Key Responsibilities for this role include the following: