AI Engineer – NoScope

TryHackMe is a rapidly expanding online platform for cybersecurity training, boasting over 6 million users and utilized by more than 1000 organizations, including government agencies and Fortune 500 companies such as Google and Microsoft.

With an initial budget of $1M, we are developing a fully autonomous AI pentesting agent capable of planning, exploiting, adapting, and reporting with the agility and accuracy required for contemporary offensive security. We are seeking an engineer to join our talented team to test and enhance our agent.

You should possess a keen interest in cybersecurity, a deep curiosity, and a passion for AI agents.

The Role

You will design the core logic of the system, influencing its reasoning, decision-making paths, memory, and execution flow to reliably handle complex offensive tasks. Your daily activities will involve close collaboration with an offensive security researcher to align the agent with actual attacker workflows and enhance its ability to identify and exploit vulnerabilities.

Responsibilities

• Contribute to the development and optimization of our autonomous AI pentesting agent.
• Build and maintain secure environments for running, testing, and benchmarking agent behaviors.
• Implement essential agent capabilities such as reasoning, decision-making, planning, and tool orchestration.
• Assist in evaluating and comparing large language models (Claude, OpenAI models, Mistral, Llama, etc.) for agent tasks.
• Develop UI components and dashboards using React and support browser automation workflows with Playwright for agent evaluation.
• Aid in the continuous refinement of the agent through experimentation, observability, and lab testing.

Requirements

• 2+ years of software development experience with strong Python skills.
• Experience in building AI agents (LangChain, CrewAI, Strands SDK, etc.).
• Hands-on experience with agent design: reasoning, memory, tool orchestration, structured outputs.
• Knowledge of prompt engineering, RAG, chain-of-thought, few-shot learning, and agent evaluation.
• Proficiency in SQL/NoSQL databases and basic data modeling.
• Familiarity with Docker, AWS, cloud deployment, and shell scripting.
• Experience with React for frontend/dashboards.

Nice to Have

• Familiarity with OWASP Top 10.
• Model training and fine-tuning (LoRA, PEFT) with evaluation experience.
• Cybersecurity expertise: OWASP attacks, pentesting methodologies.
• Experience with TryHackMe or other CTF platforms.
• Proficiency in Playwright for browser automation in agent workflows.

What We Offer

• Competitive salary.
• Equity in the AI pentesting venture.
• An early-stage role with genuine ownership and autonomy.

Please note that we currently cannot provide sponsorship.

TryHackMe simplifies the process of learning and teaching cybersecurity. Our platform offers a seamless experience through prebuilt courses that include cloud-hosted virtual machines (VMs) ready for deployment, eliminating the hassle of downloading and configuring VMs. Our platform is ideal for CTFs, workshops, assessments, or training. Since our launch in 2018, we have made significant progress, now serving over six million users!