Job DescriptionWe are seeking an experienced Senior AI Engineer specializing in AI security to join our Agentic AI team as we scale our agentic capabilities across all levels of the U.S. government.
Over the past year, we have seen rapid adoption of our AI agent, Ace. As agents gain access to increasingly powerful tools, data, and workflows, securing these systems presents a fundamentally different set of challenges from securing traditional software.
AI security is not a solved problem. This role sits at the intersection of applied AI research, offensive security, and production systems engineering. You will identify how agentic systems can fail or be exploited, develop new approaches for detecting and mitigating those failures, and build the infrastructure necessary to deploy capable AI agents securely in adversarial environments.
You will work directly with the engineers building our agent runtime, evaluation infrastructure, tools, and production AI systems. The goal is not simply to identify vulnerabilities - it is to turn what we learn into durable security architecture, automated evaluations, and engineering primitives that make our entire AI platform more secure.
This role is a full-time position located out of our office in Pittsburgh, PA or Arlington, VA and Remote for those outside of those cities.
This role may require up to 25% travel
Scope of Responsibilities- Research and develop new approaches to AI red teaming, adversarial testing, security evaluation, and robust inference.
- Threat model agentic AI architectures, identifying trust boundaries, attack surfaces, privileged capabilities, and potential failure modes.
- Design adversarial evaluations targeting threats such as prompt injection, indirect prompt injection, tool abuse, privilege escalation, data exfiltration, context or memory poisoning, and unintended agent behavior.
- Build automated security evaluation and regression frameworks that continuously test agents, models, tools, and infrastructure against known and emerging attacks.
- Translate successful attacks and research findings into production mitigations, architectural improvements, and reusable security controls.
- Design secure execution environments for AI agents interacting with tools, code, data, and external systems.
- Build and harden sandboxing and isolation mechanisms for executing agent-generated or otherwise untrusted workloads.
- Design capability boundaries, permission models, and least-privilege access controls for agent tools and services.
- Develop scalable AI infrastructure and services supporting secure model inference and agent execution.
- Own and improve production infrastructure across Kubernetes, AWS, networking, storage, and compute.
- Implement security controls around identity and access management, secrets, network isolation, containers, and service-to-service communication.
- Build scalable APIs, internal platform services, and infrastructure tooling that improve developer productivity, system reliability, and security.
- Improve observability across AI systems through structured logging, metrics, distributed tracing, dashboards, security telemetry, and automated alerting.
- Investigate complex production and security failures across models, agents, distributed systems, and infrastructure.
- Optimize performance, latency, and infrastructure cost while maintaining strong reliability and security guarantees.
- Stay current with emerging attacks against LLMs and agentic systems and rapidly translate relevant research into practical evaluations and defenses.
Qualifications- U.S. Citizenship is required
Required Skills: - 5+ years of experience building production software, backend infrastructure, distributed systems, security systems, or AI/ML infrastructure.
- Bachelor's, Master's, or Doctorate in Computer Science, Computer Engineering, Cybersecurity, Data Science, or a related technical field, or equivalent practical experience.
- Demonstrated experience in AI red teaming, adversarial machine learning, offensive security, systems security, or related security research.
- Experience evaluating AI systems beyond basic direct prompt injection attacks.
- Strong understanding of how modern LLM and agentic systems operate, including model inference, context management, tool use, retrieval, and multi-step agent execution.
- Strong intuition for how AI systems can fail when exposed to adversarial users, untrusted data, external tools, and complex production environments.
- Experience threat modeling complex systems and translating identified risks into concrete engineering controls.
- Strong programming experience in Python and experience building production-quality software.
- Experience operating production services on Kubernetes and cloud platforms such as AWS, GCP, or Azure.
- Strong understanding of networking, distributed systems, containers, service orchestration, and scalable architectures.
- Experience designing APIs, services, asynchronous systems, and event-driven architectures.
- Comfortable debugging failures that span application code, AI models, distributed systems, and infrastructure.
- Able to move between research and engineering: reading new research, developing an attack or defense, validating it experimentally, and turning the result into a production system.
- Passionate about building secure AI systems and proactively defending against novel attack vectors.
Desired Skills: - Experience designing or securing AI agent runtimes and tool-execution environments.
- Experience building secure code execution sandboxes, container isolation, microVMs, or other mechanisms for running untrusted workloads.
- Experience with cloud security, infrastructure hardening, IAM, secrets management, network isolation, and zero-trust architectures.
- Experience with offensive security, penetration testing, vulnerability research, or exploit development.
- Experience developing automated adversarial evaluations or integrating security evaluations into CI/CD pipelines.
- Experience with adversarial machine learning, model robustness, or inference-time defenses.
- Familiarity with AI security frameworks and threat taxonomies such as MITRE ATLAS, OWASP guidance for LLM/GenAI applications, or the NIST AI Risk Management Framework.
- Experience securing RAG systems, vector stores, model gateways, or other components of modern AI infrastructure.
- Experience with software supply-chain security and securing model, dependency, and container artifacts.
- Experience working in government, defense, or other high-security environments.
We firmly believe that past performance is the best indicator of future performance. If you thrive while building solutions to complex problems, are a self-starter, and are passionate about making an impact in global security, we're eager to hear from you.
