Design and implement application security test environments focused on OWASP risks in LLM applications.
Develop deterministic test cases and evaluation frameworks to distinguish vulnerable from secure code.
Collaborate with security specialists remotely to refine testing methodologies and advance AI-powered security analysis.
The company specializes in application security, AI security, and building advanced security evaluation tools. It operates as a fully remote, technically driven team that values collaboration, precision, and continuous improvement.
Design, build, and maintain cybersecurity benchmarks for AI within evaluation infrastructure using Inspect.
Sandbox and harden evaluation environments to prevent AI escape and maintain security.
Integrate with AI providers to run benchmarks and collaborate with researchers for accurate, insightful outputs.
Epoch AI investigates trends in machine learning and the economic consequences of AI, developing a comprehensive knowledge base for policymakers. The research institute is committed to inclusive, equitable community and produces rigorous, accessible work on compute trends, data scarcity, and algorithmic progress.
Design and build red team agents, architecting harnesses, tools, and MCP servers to encode offensive tradecraft.
Drive full-chain attack operations across Windows, Active Directory, on-prem, cloud, and network surfaces.
Build detection-evasion capabilities, cyber ranges, and benchmarks to verify agent improvements against real defenses.
Cogent is an applied AI lab building AI agents for cybersecurity, using its AI Taskforce to assess petabytes of enterprise data and prevent breaches. Backed by Greylock and Bain Capital Ventures, the team includes experts from top universities, leading tech companies, and world-class cybersecurity firms.
Apply advanced mechanical engineering expertise to create, solve, and validate engineering tasks for AI training.
Use computational modeling, simulation, and AI coding agents to develop and assess engineering solutions.
Work fully remotely as an independent contractor on a 12-month contract, contributing to AI system reliability.
We are an AI-powered job matching platform that connects top candidates with hiring companies. We use AI to review applications and shortlist the best-fitting candidates for a fair and efficient recruitment process.
Define the long-term technical vision for AI/LLM offense and drive its integration into NodeZero's autonomous pentesting platform.
Lead research and productization of attacks on LLM applications, agentic systems, and AI infrastructure.
Partner with senior leadership to shape strategy, standards, and execution across engineering teams.
Horizon3 is a fast-growing, remote cybersecurity company building NodeZero, an autonomous pentesting platform. The team is a fusion of Special Operations cyber operators and startup engineers, with a culture of respect, collaboration, and results.
Design and build automated evaluation pipelines for AI agents.
Develop testing frameworks for reasoning, multi-turn conversations, and adversarial scenarios.
Own the evaluation architecture and integrate observability tools in a fully remote team.
A company building production-grade AI agents and systems. They operate in a fully remote, globally distributed environment with a culture focused on innovation, autonomy, and continuous improvement.
Conduct original offensive research against Dia, including prompt injection, sandbox escape, and cross-profile data access.
Threat model new surfaces with engineering teams and review features before they ship.
Build automated vulnerability discovery systems, including fuzzing and AI-assisted hunting pipelines.
The Browser Company is building a better way to use the internet, with a focus on an AI-powered browser called Dia. They are a remote-first team of nearly 100 people who value diversity and craft.
Conduct manual pentests on core systems and AI agents, building automation to expand coverage.
Define how we pentest AI, including LLM applications and agent-generated code.
Triage SAST findings, review code across web/APIs/AWS, and communicate risk to engineers.
We are a financial technology company based in Boston, unlocking capital for small businesses across America with fast, flexible funding. Our remote-first team, recognized as a Best Place to Work, has provided over $4.8 billion to more than 92,000 small businesses since 2012.
Review existing RL environments for offensive-security and application-security tasks, confirming vulnerabilities are real, reachable, and solvable with provided context.
For patching and exploit-construction tasks, trace or attempt the intended path and verify grading correctly distinguishes genuine fixes from pattern-matching or workarounds.
Flag inconsistent files, missing context, or ambiguity, and write clear justifications for each accept, reject, or flag decision.
idlexperts is idler's quality assurance network of part-time, remote experts who review reinforcement-learning environments before they reach customers. The culture is flexible, async, and self-directed, with experts expected to work a handful of hours per week and deliver clear written judgments.
Review defensive-security and infrastructure-security RL environments for realism, solvability, and correct grading.
Flag missing context, inconsistent files, unclear requirements, or anything that makes tasks ambiguous.
Write concise justifications for each decision and work async on a flexible part-time schedule.
IdleXperts is idler's quality assurance network, made up of part-time remote experts who review reinforcement-learning environments before customer delivery. The company relies on independent, self-directed contributors with deep security expertise to ensure tasks are fair and correctly graded.
Lead technical discovery and turn ambiguous cyber needs into a tractable plan with clear architecture and acceptance criteria.
Build and integrate AI-enabled security workflows using production software, APIs, telemetry, and security systems.
Deploy, operate, and evaluate AI workflows in customer environments, ensuring security, maintainability, and handover.
Eliza is a technology services company and Advanced-tier OpenAI partner that helps organizations build and deploy AI solutions, including generative AI, custom LLM integrations, predictive analytics, and intelligent automation. The company has a collaborative, mission-driven team focused on real-world AI impact and values adaptability, ownership, and hands-on client collaboration.
Design evaluations of conceptual reasoning in AI safety, covering governance, alignment, macrostrategy, and economics.
Develop rubrics that separate genuine reasoning from plausible-sounding filler.
Iterate on tasks with frontier models and collaborate with a top research team.
Trajectory Labs is a startup automating AI safety research. We build safety and alignment evals, red-teaming programs, and RL environments for frontier AI labs. As a one-year-old company, we've been featured in major frontier lab safety reports and led high-impact red-teaming efforts.
Run full kill chains against web, mobile, API, and thick/thin clients, proving exploitability and mapping blast radius.
Train the swarm by translating your tradecraft into reasoning and evaluation data for autonomous AI attackers.
Pressure-test the swarm, build tools, and tell the attack story to engineers and executives.
Armadin builds autonomous proactive security, deploying AI agents that discover and remediate threats before breaches. We're a scrappy, high-ownership team of elite researchers and engineers, led by Kevin Mandia, with startup-level autonomy.
Work across strategic customer accounts to make Dropzone more trusted and embedded in daily security operations.
Partner with TAMs, Product, and Engineering to review investigations, tune configurations, and refine response workflows.
Translate customer SOC practices into Dropzone behavior, measure impact, and build repeatable playbooks.
Dropzone AI is a venture-backed cybersecurity company using Gen AI to offload repetitive work for security analysts and transform cyber defense. The team has deep experience across cybersecurity, AI/ML, and SaaS product development, with an early-stage, high-ownership remote culture.
Own technical security controls for AI usage, including agent identity, tool access, guardrails, and logging.
Partner with AI Enablement, Engineering, DevOps, GRC, and Security to integrate security by design.
Build AI-powered security tooling and lead AI-specific incident response.
Alpaca is a US-headquartered global leader in agent-first brokerage infrastructure, serving financial institutions across 40 countries. With 400+ distributed team members and $400M in funding, we are a diverse, remote-first team opening financial services to everyone.
Work through complex mechanical engineering problems while narrating your reasoning process.
Evaluate AI-generated responses for accuracy, depth, and reasoning quality.
Write improved exemplars and structured feedback to enhance AI model performance.
Prolific is building the largest pool of quality human data for AI, used by over 35,000 developers and researchers. We provide a global platform for flexible, remote work to contribute to ethical AI development.
Lead AI-powered automation initiatives across client-facing cybersecurity services.
Combine software engineering, AI, and security expertise to design scalable solutions.
Mentor multidisciplinary teams and collaborate with senior stakeholders on technical strategy.
The company provides client-facing cybersecurity services, leveraging artificial intelligence to automate and enhance security operations. It fosters a collaborative, multidisciplinary culture focused on continuous improvement and service innovation.
Deliver end-to-end threat intelligence enablement and technical onboarding, optimizing customer digital footprints and alert rules for noise-free monitoring.
Develop custom automation scripts, AI agents, and SOAR playbooks to streamline security workflows and enhance platform adoption.
Drive customer retention through proactive engagement, churn risk analysis, and adaptive threat reporting.
SOCRadar provides Extended Threat Intelligence (XTI) solutions that help organizations proactively secure their attack surfaces and gain visibility into surface, deep, and dark web threats. The company is focused on innovation, global expansion, and delivering automation-enabled security to enterprises worldwide.