Agentic AI Security Engineer

View email address on us.fitly.work at Janelia integrates AI agents across the research process. These tools write and execute code, optimize experimental and model parameters, and drive instruments.

Letting AI agents act on shared research infrastructure requires serious safety engineering. The relevant failure mode is usually not a malicious attacker but a capable, misdirected tool running with legitimate user permissions at machine speed: hallucinated commands, instructions hijacked by retrieved content, runaway loops, sandbox escapes. We are hiring the engineer who owns this problem.
The role You will research, develop, deploy, operate, and analyze the safety stack for agentic AI at Janelia:

• Sandboxing. Hardened, reproducible execution environments for agent code (kernel- and VM-level isolation, network egress control) on our OpenShift platform and HPC cluster, validated by adversarial testing.
• Guardrails. Policy enforcement on agent inputs, outputs, and tool calls; judge models; human approval gates; independent safeguard services that sessions can be routed through, individually or chained.
• Observability. Attributable audit logging, full-trace capture, real-time monitoring, and post-incident analysis of agent sessions.

You will join early enough to make the foundational architecture decisions, and you will operate what you design. We expect and support open-sourcing and publishing this work.
The environment A large NVIDIA GPU cluster (B300/H200/H100 class) running self-hosted open-weight models and frontier commercial models; collaborations with industry (Anthropic and Google); scientists across biology, physics, and ML as your users; the long-term stability of a philanthropy-funded research institute.
What we look for

• Hands-on experience operating LLM agents and dealing with their failures, or deep sandbox/container-security expertise with clear ability to move into the agentic domain.
• Linux and Kubernetes/OpenShift isolation depth, including a precise understanding of the difference between containerization and sandboxing.
• Working knowledge of agent architectures: tool-calling protocols (e.g., MCP), approval gates, judge models, guardrail frameworks, and the limits of each.
• Evidence over certifications: open-source contributions, publications or technical writing, disclosed findings, or production systems you can discuss in depth.
• Minimum 5 years in security, systems/SRE, or ML systems engineering.

To apply: include a link to something you have built, broken, or written.

Physical Requirements

Remaining in a normal seated or standing position for extended periods of time; reaching and grasping by extending hand(s) or arm(s); dexterity to manipulate objects with fingers, for example using a keyboard; communication skills using the spoken word; ability to see and hear within normal parameters; ability to move about workspace. The position requires mobility, including the ability to move materials weighing up to several pounds (such as a laptop computer or tablet).

Persons with disabilities may be able to perform the essential duties of this position with reasonable accommodation. Requests for reasonable accommodation will be evaluated on an individual basis.

Please Note:

This job description sets forth the job’s principal duties, responsibilities, and requirements; it should not be construed as an exhaustive statement, however. Unless they begin with the word “may,” the Essential Duties and Responsibilities described above are “essential functions” of the job, as defined by the Americans with Disabilities Act.

#LI-BG1

Hiring Pay Range

Agentic AI Security Engineer: $149,084.80 - $186,356.00Agentic AI Security Senior Engineer: $181,143.20 - $226,429.00Agentic AI Security Principal Engineer: $210,803.20 - $263,504.00
Pay Type:Salary
The posted range reflects HHMI’s good faith estimate of the anticipated hiring salary range for this role at the time of posting. Actual hiring compensation is determined by a candidate’s qualifications, experience, and internal equity.