Material works with some of the most sophisticated security teams in the industry. These customers have entrusted us with protecting one of their most critical assets: Google Workspace. Your mission is to make our own corporate environment worthy of that trust: take a mature security program and make it rigorous, measured, and adapted to the new generation of LLM-based security tooling. Success means we can detect and defend against sophisticated attacks. This includes hardening our corp services, increasing detection efficacy while reducing response times, and making sure our employees are working with customer data in ways that reflect the trust of our customers Concretely: bring your experience to identify and close gaps in our environment, rebuild our detection and response around LLMs — and be customer zero for our own product, working with Engineering to turn what you learn into detection improvements that ship to every customer. The environment is what you’d expect at a modern startup: Mac-only fleet, Google Workspace, expected security services like a SIEM, IdP, EDR, fleet management and a cloud production environment. Our security posture is more mature than typical for a company our size.
You Might Be This Person If
- You have owned a corporate stack end to end — devices, identity, SaaS, collaboration — and turned controls that were deployed into controls that are enforced.
- You treat the IdP, the device fleet, and the SaaS estate as one system, and you harden the paths between them rather than each tool in isolation.
- You have written detections that caught real attacks and removed the noise that buried the team, and you can explain how you measured the difference.
- You have responded to live incidents and made the containment call. You know what to revoke first.
- When a job requires someone to do something manually and repetitively, your instinct is to build the system that makes that job unnecessary.
- You have put an LLM into a security workflow, measured it, and have specific views about where it should and should not be trusted.
- You work through design decisions by asking questions on both sides, and when the call goes against you, you respect it and build it.
- Bonus: you have written a technical paper, spoken at a technical conference, or maintain open-source security tooling.
Responsibilities
- Detection and Response (own): Own corporate telemetry and the detection content on it. Close coverage gaps, reduce noise with measured fidelity, and build detections for what matters here: account takeover, session token theft, malicious OAuth grants, admin abuse, and exfiltration by someone who legitimately has access. Lead corporate incident response and own the playbooks.
- LLM-Driven Triage (own): Design and build the agent that reviews privileged access grants, customer data access, and high-fidelity endpoint and identity alerts in near real time, with labeled evaluation sets, measured precision and recall, explicit limits on automated action, and a defense against being steered by log content or an employee's reply.
- Endpoint (shared with the CorpSec team): Device management and endpoint detection for the Mac fleet. Automate update enforcement, deepen EDR coverage, bring browser-level DLP to the fleet, and enforce BYOD posture on iOS and Android.
- Identity and SaaS (shared with the CorpSec team): The identity provider and the SaaS estate. Maintain phishing-resistant authentication everywhere, address the applications that cannot do SSO, automate access-review evidence, and turn manual vendor and OAuth application review into a pipeline.
- Google Workspace (shared with the CorpSec team): Admin alerting, OAuth governance, Apps Script governance, sharing, and DLP, using our own product where it fits. Partner with Engineering to propose, test, and ship detection updates to the product based on what you see in our own tenant.
- Corporate to Production (shared with AppSec): The boundary between corporate and production: detections that cross it, just-in-time access, and the logging that shows who touched what.
What We’re Looking For
- Operator Depth: 6+ years in corporate or enterprise security or detection and response, with hands-on ownership of a SIEM, detection content, and incident response.
- Software Engineering: Comfortable in Python, Go, or TypeScript working directly against vendor APIs. You automate enforcement rather than configuring it by hand.
- Identity Fundamentals: Deep understanding of OAuth, SAML/OIDC, session and token lifecycle, and how identity reaches cloud IAM.
- Endpoint Fluency: You have managed a Mac fleet with a modern MDM and EDR and taken it from deployed to enforced.
- Applied LLM Judgment: You have built or seriously experimented with LLM-driven workflows, including prompting, tool use, and evaluation, and can explain how you would know whether one is doing a good job.
- Software Development (nice to have): You have contributed detection logic or security features to a product codebase and are comfortable working inside an engineering team's review and release process.
- How You Work: Collaborative, curious, direct. You work well next to more junior engineers without taking over, and you would rather engineer a problem away than staff it.
-- Material Security is a remote-first workplace with an office in San Francisco, California. By clicking "Apply for this Job", you acknowledge that you have read the California Candidate Privacy Notice Regarding Use of Personal Information and hereby agree to its terms. Compensation at Material Security is determined by a range of factors, including but not limited to the individual’s particular combination of knowledge, skills, competencies, and experience. The projected compensation range for this position is $220,000-$240,000 annually.
Equal Opportunity Employer Statement Material Security is an equal opportunity employer. We celebrate diversity and are committed to creatin
Sourced from the a16z Speedrun talent network — apply on Speedrun.