Security Data Engineer

Anthropic•San Francisco, CA

7h•$320,000 - $405,000•Hybrid

About The Position

Anthropic is hiring a Security Data Engineer to own the data and systems used to secure the data center footprint where Claude is trained and served. This includes the sites, hardware, and people and vendors that handle it. The role owns component-level hardware inventory and chain of custody, software that captures site and host activity, continuous control monitoring, and calculations that determine what each site is cleared to run. The engineer will set architecture, agree on schemas and interfaces with source-system owners, build data pipelines, stores, tools, and services, and operate what they ship. The goal is to continuously verify custody, access, and evidence using reliable records, tools, and AI agents. The role works closely with DCSE security monitoring engineers, compute and data center operations teams, data center operators, hardware vendors, and logistics providers. The position is based in San Francisco, with US remote applicants considered, and requires regular travel to data center sites.

Requirements

Experience designing and operating production data systems, including relational schemas, change data capture, batch or streaming pipelines, and supporting services.
Experience maintaining data correctness through measured coverage, reconciliation against sources, alerts when feeds go silent, and designs that account for corrupted or lost records.
Experience integrating with systems outside your control, such as inventory tools, ticketing systems, or vendor APIs, and building data models that are not tied to a single source.
Experience shipping and maintaining software outside a central service, such as host agents, site-local services, or edge collectors, or building metrics used and defended in team decisions.
Hands-on experience with data about physical assets or security controls, such as a hardware fleet, data center or supply-chain inventory, asset custody, or security and compliance control evidence.
Ability to read and review others’ code and build with AI coding agents using Git and code review.
Ability to own a domain end to end, including setting the approach, designing and building with dependent teams and vendors, and operating the resulting system.
Strong SQL skills and proficiency in at least one general-purpose programming language.
Bachelor’s degree or an equivalent combination of education, training, and/or experience.
A field of study relevant to the role, demonstrated through coursework, training, or professional experience.

Nice To Haves

Experience with asset management, supply-chain, or warehouse systems where serialized tracking and custody transfer are central.
Experience building continuous control monitoring or security or compliance program metrics, or producing evidence accepted by an assessor, such as against NIST SP 800-53, ISO 27001, or SOC 2.
Experience with risk or posture scoring that combines control results, findings, and other evidence into a decision-making measure.
Familiarity with data center hardware lifecycles from receipt to decommissioning and with external sites and parties, including colocation, cloud or GPU cloud operators, hardware manufacturers, and logistics providers.
Experience running software across many hosts or remote sites with versioned rollout and a way to verify each copy is still reporting.
Experience building AI agents or automation that replaces recurring manual work.

Responsibilities

Own the component-level hardware inventory and chain-of-custody record, including its data model, pipelines joining inventory and custody events from multiple source systems, record change history, and tamper-evident stores.
Build tools and AI agents for reconciliation, audit-evidence exports, and custody checks, replacing recurring manual review work where possible.
Build custody workflows and site and host software that capture events such as receipts, moves, data wipes, and decommissioning, and enforce authorization to record them.
Own pipelines that turn check outcomes, assessment results, and findings into per-control and per-site measures over time, and maintain the posture calculation determining what each site is cleared to run.
Build data-coverage checks, reconciliation against source systems, and alerts for silent or drifting feeds; ensure incorrect or missing records are detected promptly and own retention and access controls.
Agree on schemas and interfaces with security monitoring platform engineers and build query surfaces and reports for the team, site leads, and leadership.
Set the approach for a domain, design and build it with dependent teams and vendors, and operate it as a system people use every day.