5 hours ago
Base Salary
$210k - $234k/yr
Responsibilities
- Own the AI product security roadmap, including the multi-year vision, strategy, and standards for securing Compass AI capabilities.
- Threat model AI features, RAG pipelines, multi-agent workflows, and automated decisioning before release.
- Architect and automate real-time protections for LLM inputs and outputs, including injection mitigation, jailbreak defense, output verification, and sensitive-data egress checks.
- Design controls for autonomous agents, including scoped permissions, human-in-the-loop checkpoints, audit trails, and safe tool invocation.
- Secure retrieval systems through access controls, source integrity, tenancy boundaries, and retrieval-boundary protections.
- Assess the security, privacy, retention, and training-use posture of third-party models, inference providers, and vector stores.
- Build telemetry and detections for AI abuse, anomalous tool invocation, prompt abuse, unusual retrieval, and model-mediated data exfiltration.
- Ship reusable security libraries, CI/CD checks, and evaluation harnesses that make secure AI patterns the default.
- Advise engineers and executives on AI architecture, emerging risks, and security roadmaps.
Requirements
- 7+ years of security engineering experience, including 2+ years focused on AI/ML or LLM application security.
- Demonstrated experience securing a production AI or ML system across development, deployment, and runtime.
- Strong proficiency in Python or Go and experience building production integrations and security pipelines.
- Deep working knowledge of transformer architectures, inference pipelines, RAG, prompt engineering, and the OWASP Top 10 for LLMs.
- Hands-on experience with adversarial ML, prompt injection, model poisoning, data extraction, unsafe tool use, and red teaming.
- Strong application and cloud security fundamentals, including AWS and/or GCP, Kubernetes, IAM, API security, and DevSecOps practices.
- Experience threat modeling complex distributed systems and driving remediation with engineering teams.
- Strong strategic communication skills and the ability to translate AI risk into technical and business impact.
- Experience with multi-agent autonomous systems or Model Context Protocol architectures is preferred.
- Familiarity with MITRE ATLAS, NIST AI RMF, or comparable AI threat and risk frameworks is preferred.
- Experience building LLM evaluation and safety-testing harnesses, securing AI platforms or SDKs, or working in regulated or consumer-data-heavy environments is preferred.
Benefits
- Base pay range is $210,000-$234,000, with possible bonuses and restricted stock units in addition to benefits.
- Paid vacation, holidays, sick time, parental leave, and recharge leave.
- Medical, tele-health, dental, vision, 401(k), flexible spending accounts, commuter program, life and disability insurance.
- Maven new-parent support, Carrot fertility benefits, UrbanSitter caregiver referrals, Employee Assistance Program, and pet insurance.
- Participation in incentive programs that may include eligible cash, equity, or commissions.
Categories
About Compass
Compass is a residential real estate brokerage and technology company that builds an end-to-end platform for agents and their clients. It earns primarily from brokerage commissions while providing CRM, marketing, listings, and transaction tools to help agents serve home buyers and sellers. Founded in 2012 and headquartered in New York City, Compass operates across major U.S. markets and is publicly traded on the NYSE.