Founding AI Engineer, Offensive Security
We are hiring a Founding AI Engineer to help lead RemoteThreat’s AI strategy, architecture, and engineering direction across the platform. This is a role for someone who understands not only how modern AI systems behave, but how to design, build, evaluate, and operationalize them in demanding real-world environments.
You will define how AI is architected inside RemoteThreat: how models, agents, tools, memory, orchestration, evaluation, and user-facing workflows fit together to create durable advantage for operators. You will work across research, engineering, product, and leadership to build AI-native capabilities that improve mission speed, workflow quality, decision support, task decomposition, and continuous adversary simulation.
This is not a pure research role and it is not a generic innovation role. We are looking for someone who can set technical direction for AI systems end to end, from model selection and agent design to backend architecture, evaluation pipelines, and production-grade integration into security workflows.
What You’ll Do
Own AI strategy and technical direction across RemoteThreat, including architecture, research priorities, capability roadmap, evaluation standards, and production integration.
Design the core AI architecture for the platform, including model routing, agent orchestration, tool use, memory patterns, retrieval layers, context management, and security boundaries.
Lead development of AI-enabled capabilities across the platform, including operator assistance, reasoning support, workflow automation, mission planning support, and knowledge capture.
Define how AI systems should interact with the rest of the RemoteThreat platform, including exposure discovery, adversary simulation, detection engineering, and team training workflows.
Establish engineering standards for building reliable AI systems, including observability, prompt and context versioning, evaluation harnesses, latency/performance tradeoffs, fallback behavior, and human-in-the-loop controls.
Drive internal research into the strengths and limits of state-of-the-art models across tasks such as vulnerability discovery, tool use, multi-step reasoning, chained operations, and security-relevant decision support.
Build repeatable evaluation systems for model quality, operational usefulness, misuse resistance, reliability, and failure-mode discovery in offensive and adversarial workflows.
Partner closely with engineering to turn prototypes into robust product capabilities rather than one-off demos or disconnected experiments.
Work with offensive operators and customer-facing teams to ensure the architecture reflects actual mission constraints, not just idealized AI workflows.
Build and lead a small, elite AI engineering and research function capable of shipping differentiated systems quickly and rigorously.
Must Have
Deep experience leading applied AI, LLM, or agent-based initiatives in security, cyber, safety, or similarly technical domains.
Strong architectural judgment for AI systems, including experience designing production-grade workflows involving models, agents, tools, retrieval, memory, orchestration, and evaluation.
Experience building or overseeing the engineering of AI-enabled platforms, not just prompting or experimentation layers.
Strong understanding of modern foundation models, agentic systems, tool use, model evaluation, prompt strategies, and the practical limits of current systems.
Experience designing evaluation pipelines for model behavior in adversarial, ambiguous, or high-consequence settings.
Strong technical fluency across at least some of the following: offensive security, red teaming, vulnerability research, AI security, agent systems, model benchmarking, backend systems, or workflow automation.
Experience making hard system tradeoffs across speed, cost, reliability, transparency, and operator trust.
Proven ability to set technical direction, mentor senior engineers and researchers, and align research with product and platform outcomes.
Excellent written and verbal communication skills, especially around architecture decisions, technical tradeoffs, and executive-level technical strategy.
Citizenship in a NATO country, or a citizen of Taiwan, Australia, New Zealand, South Korea, Japan, UAE, Israel, Latin America, South America (excluding Venezuela)
Strongly Preferred
Experience building agent frameworks or AI orchestration layers for complex technical tasks.
Experience with production AI infrastructure, including inference patterns, model gateways, evaluation services, telemetry, and backend-heavy AI systems.
Experience evaluating or building AI systems for offensive cyber, red teaming, security testing, or capability assessment.
Familiarity with model behavior in areas such as tool use, multi-turn attack flows, prompt injection, sandbox escape risk, agent reliability, and operational misuse.
Experience developing internal benchmarks, taxonomies, or structured evaluation methods for cyber capability, safety, or operational uplift.
Hands-on coding fluency in Python and comfort working across notebooks, services, APIs, and engineering-heavy research environments.
Experience working in high-trust, mission-oriented, or sensitive technical environments where reliability and discretion matter.
Environment You Will Work In
AI-native systems embedded in offensive cyber and adversary simulation workflows.
Architectures involving model routing, tool invocation, agent loops, retrieval layers, memory management, and workflow instrumentation.
Evaluation pipelines designed to measure not only benchmark performance, but real-world utility, robustness, and failure behavior.
Backend-heavy platforms where orchestration, logging, observability, security, and reliability are as important as model quality.
Sensitive enterprise and mission environments where AI systems must perform under operational constraints rather than ideal lab conditions.
What Success Looks Like
RemoteThreat develops a differentiated AI architecture that is purpose-built for offensive cyber and adversary simulation rather than copied from generic enterprise AI patterns.
AI systems across the platform are reliable, measurable, and deeply integrated into real operator workflows.
The company builds a strong internal capability around AI engineering, orchestration, and evaluation, not just experimentation.
RemoteThreat is able to ship AI features that improve operator speed, reasoning quality, and scale without sacrificing trust, control, or technical rigor.
AI becomes a durable part of the platform’s technical advantage and not a thin layer on top of existing products.
Benefits
Fully Remote
Competitive salary and equity stock options
Unlimited paid time off (PTO)
Company-paid holidays
Employer subsidized medical/vision and dental coverage
HSA and FSA account options
24/7 Virtual medical and behavioral health consultations
Company paid mental healthcare sessions
Eligibility to participate in the company’s incentive bonus program
Company paid life insurance
Company paid short- and long-term disability
Travel assistance program (emergency assistance, medical services, identity protection)
Parental leave
Employee discount program
Financial well-being platform
Unlimited AI token usage
You will define how AI is architected inside RemoteThreat: how models, agents, tools, memory, orchestration, evaluation, and user-facing workflows fit together to create durable advantage for operators. You will work across research, engineering, product, and leadership to build AI-native capabilities that improve mission speed, workflow quality, decision support, task decomposition, and continuous adversary simulation.
This is not a pure research role and it is not a generic innovation role. We are looking for someone who can set technical direction for AI systems end to end, from model selection and agent design to backend architecture, evaluation pipelines, and production-grade integration into security workflows.
What You’ll Do
Own AI strategy and technical direction across RemoteThreat, including architecture, research priorities, capability roadmap, evaluation standards, and production integration.
Design the core AI architecture for the platform, including model routing, agent orchestration, tool use, memory patterns, retrieval layers, context management, and security boundaries.
Lead development of AI-enabled capabilities across the platform, including operator assistance, reasoning support, workflow automation, mission planning support, and knowledge capture.
Define how AI systems should interact with the rest of the RemoteThreat platform, including exposure discovery, adversary simulation, detection engineering, and team training workflows.
Establish engineering standards for building reliable AI systems, including observability, prompt and context versioning, evaluation harnesses, latency/performance tradeoffs, fallback behavior, and human-in-the-loop controls.
Drive internal research into the strengths and limits of state-of-the-art models across tasks such as vulnerability discovery, tool use, multi-step reasoning, chained operations, and security-relevant decision support.
Build repeatable evaluation systems for model quality, operational usefulness, misuse resistance, reliability, and failure-mode discovery in offensive and adversarial workflows.
Partner closely with engineering to turn prototypes into robust product capabilities rather than one-off demos or disconnected experiments.
Work with offensive operators and customer-facing teams to ensure the architecture reflects actual mission constraints, not just idealized AI workflows.
Build and lead a small, elite AI engineering and research function capable of shipping differentiated systems quickly and rigorously.
Must Have
Deep experience leading applied AI, LLM, or agent-based initiatives in security, cyber, safety, or similarly technical domains.
Strong architectural judgment for AI systems, including experience designing production-grade workflows involving models, agents, tools, retrieval, memory, orchestration, and evaluation.
Experience building or overseeing the engineering of AI-enabled platforms, not just prompting or experimentation layers.
Strong understanding of modern foundation models, agentic systems, tool use, model evaluation, prompt strategies, and the practical limits of current systems.
Experience designing evaluation pipelines for model behavior in adversarial, ambiguous, or high-consequence settings.
Strong technical fluency across at least some of the following: offensive security, red teaming, vulnerability research, AI security, agent systems, model benchmarking, backend systems, or workflow automation.
Experience making hard system tradeoffs across speed, cost, reliability, transparency, and operator trust.
Proven ability to set technical direction, mentor senior engineers and researchers, and align research with product and platform outcomes.
Excellent written and verbal communication skills, especially around architecture decisions, technical tradeoffs, and executive-level technical strategy.
Citizenship in a NATO country, or a citizen of Taiwan, Australia, New Zealand, South Korea, Japan, UAE, Israel, Latin America, South America (excluding Venezuela)
Strongly Preferred
Experience building agent frameworks or AI orchestration layers for complex technical tasks.
Experience with production AI infrastructure, including inference patterns, model gateways, evaluation services, telemetry, and backend-heavy AI systems.
Experience evaluating or building AI systems for offensive cyber, red teaming, security testing, or capability assessment.
Familiarity with model behavior in areas such as tool use, multi-turn attack flows, prompt injection, sandbox escape risk, agent reliability, and operational misuse.
Experience developing internal benchmarks, taxonomies, or structured evaluation methods for cyber capability, safety, or operational uplift.
Hands-on coding fluency in Python and comfort working across notebooks, services, APIs, and engineering-heavy research environments.
Experience working in high-trust, mission-oriented, or sensitive technical environments where reliability and discretion matter.
Environment You Will Work In
AI-native systems embedded in offensive cyber and adversary simulation workflows.
Architectures involving model routing, tool invocation, agent loops, retrieval layers, memory management, and workflow instrumentation.
Evaluation pipelines designed to measure not only benchmark performance, but real-world utility, robustness, and failure behavior.
Backend-heavy platforms where orchestration, logging, observability, security, and reliability are as important as model quality.
Sensitive enterprise and mission environments where AI systems must perform under operational constraints rather than ideal lab conditions.
What Success Looks Like
RemoteThreat develops a differentiated AI architecture that is purpose-built for offensive cyber and adversary simulation rather than copied from generic enterprise AI patterns.
AI systems across the platform are reliable, measurable, and deeply integrated into real operator workflows.
The company builds a strong internal capability around AI engineering, orchestration, and evaluation, not just experimentation.
RemoteThreat is able to ship AI features that improve operator speed, reasoning quality, and scale without sacrificing trust, control, or technical rigor.
AI becomes a durable part of the platform’s technical advantage and not a thin layer on top of existing products.
Benefits
Fully Remote
Competitive salary and equity stock options
Unlimited paid time off (PTO)
Company-paid holidays
Employer subsidized medical/vision and dental coverage
HSA and FSA account options
24/7 Virtual medical and behavioral health consultations
Company paid mental healthcare sessions
Eligibility to participate in the company’s incentive bonus program
Company paid life insurance
Company paid short- and long-term disability
Travel assistance program (emergency assistance, medical services, identity protection)
Parental leave
Employee discount program
Financial well-being platform
Unlimited AI token usage