Job Description

Job Description:

Science Vault is currently a working but mostly manual data publishing workflow for research/scientific computing data. It publishes data from ASC/PCluster environments into an S3-backed, read-only storage model with metadata capture, access through sciencevault.amgen.com, and controls tied to Okta, AD groups, and S3 permissions.

The immediate need is someone who can own the closing engineering work: stabilize the workflow, reduce manual steps, improve usability, and help make it supportable for broader Research use.

Core skills needed:

Strong Python scripting and CLI development

AWS S3 experience, including bucket structure, versioning, permissions, and access patterns

IAM/security awareness, ideally with Okta, AD groups, or enterprise auth integration

Linux/HPC environment familiarity, especially working from shared filesystems or cluster environments

Ability to work with metadata capture, validation, and structured data manifests

Practical debugging, logging, packaging, and operational hardening experience

Helpful additional skills:

Web app or lightweight portal development for self-service publishing/access

Familiarity with data cataloguing, FAIR metadata, MMDS, or scientific data management

Experience building supportable internal tools with documentation and handoff materials

Comfort working with scientists/research users to convert rough workflows into usable products