Senior Security Engineer, AI Enablement

JLL•Los Angeles, CA

1d•$200,000 - $225,000•Remote

About The Position

The Senior Security Engineer, AI Enablement is Security's embedded, full-time representative on JLL's Falcon team, owning the product's security architecture for agents, MCP integrations, and identity end-to-end rather than reviewing it after the fact—advancing APEX + 2030's mandate that AI products are built securely by design, not secured after launch. Reporting to the Head of AI Security while sitting day-to-day with Falcon's engineers, this role requires the technical depth to write and review production code, architect agentic identity and permissioning, and represent Falcon's security posture credibly to the broader Cybersecurity organization.

Requirements

6+ years of security engineering experience, including meaningful time embedded directly within a software product or engineering team rather than in a purely advisory security function.
Strong software engineering skills, with the fluency to read, write, and review production code alongside the engineers you're partnered with.
Hands-on experience with agentic AI architectures, including MCP servers, tool-calling, and the Claude Agent SDK or a comparable agent framework.
Deep working knowledge of identity and access concepts, including OAuth/OIDC, service-to-service authentication, and least-privilege credential design for both human and non-human identities.
Demonstrated ability to operate as a technical peer to software engineers, not just a reviewer, earning influence through credibility rather than process authority.
Excellent cross-functional communication skills—able to represent a single product team's architecture accurately to the broader security organization and vice versa.

Nice To Haves

Bachelor's degree in Computer Science, Engineering, or a related field, or equivalent practical experience.
Prior experience as an embedded or product security engineer within an agile product team.
Direct experience securing multi-agent systems, RAG pipelines, or MCP-based tool integrations in production.
Familiarity with the OWASP LLM Top 10, MITRE ATLAS, or comparable agentic and AI threat frameworks.
Experience with workload identity frameworks, such as SPIFFE/SPIRE, or modern non-human identity management.
Certifications such as CISSP, OSCP, or GIAC credentials focused on application or cloud security.

Responsibilities

Sit inside the Falcon team as security's full-time technical representative, participating in planning, design reviews, and sprint work rather than reviewing finished features after the fact.
Serve as the primary point of contact between Falcon and the broader Cybersecurity organization—translating security requirements into product decisions the team can act on immediately.
Build enough trust and technical credibility with Falcon's engineers that security is treated as a design input, not a release gate.
Own the security architecture for Falcon's use of agents, MCP servers, and tool integrations, defining permission boundaries, trust boundaries, and blast radius for every agent-to-tool and agent-to-agent interaction.
Review and harden the identity model underpinning Falcon's agentic components, including: Service-to-service and agent-to-tool authentication, Scoped, least-privilege credentials for every MCP server and tool connection, Session and delegation boundaries when agents act on a user's behalf, Audit logging sufficient to reconstruct what an agent did and why.
Design secure patterns for prompt handling, tool-calling, and data flow specific to Falcon's product surface, informed by how the Claude Agent SDK and MCP actually work under the hood.
Partner with the Head of AI Security to align Falcon's architecture with enterprise AI security standards without slowing product velocity.
Own the security component of Falcon's product enablement work end-to-end, from threat modeling new features to defining the guardrails that ship with them.
Write and maintain the security requirements, configuration standards, and secure-by-default patterns that Falcon's engineers build against, rather than producing a review checklist after the fact.
Define what "secure enough to ship" means for each Falcon release—balancing genuine risk against the team's delivery timeline.
Represent Falcon's security posture and roadmap to the broader Cybersecurity organization, including the Head of AI Security and other AI Security team members.
Provide deep technical input into cross-functional security discussions, including third-party risk assessments and enterprise AI governance decisions where Falcon's architecture is relevant.
Collaborate with identity, cloud security, and application security teams to ensure Falcon's agentic components meet the same bar as any other production system.
Write and review code directly, contributing to Falcon's codebase where security-critical components such as auth, permissioning, and agent orchestration are involved, rather than advising from the sidelines.
Maintain hands-on fluency with the Claude Agent SDK, MCP, and modern identity protocols, including OAuth, OIDC, and workload identity standards, as Falcon's architecture evolves.
Stay current on agentic security research and real-world incidents—bringing relevant lessons back to Falcon before they become the team's problem.