Autonomous Security Engineer — LLM-Powered Pentest Platform

Horizon3; United States; Hybrid; USD 169,000 - 208,000; Full time

Benefits offered by this job

Inclusive team

Growth opportunities

Innovative culture

Hybrid & Remote Work

Competitive compensation

Job summary

Horizon3.ai is building an autonomous web application penetration tester that operates production-safe, autonomous attacks and scales across complex environments. We seek an engineer who can convert deep offensive expertise into reliable agent capabilities, focusing on orchestration, tooling, and verification to avoid false positives.

You’ll own the agent harness, evaluation infrastructure, and production LLM management, while ensuring deterministic behavior and scalable performance in a

Qualifications

• 5+ years building production software, with strong Python.
• Hands-on experience building LLM-powered applications or agents.
• Track record of making LLMs reliable in production.
• Experience with evaluation: harness that measures improvements.
• Strong instincts for prompt and context engineering.
• Solid software fundamentals — testing and observability.
• Ownership mentality, comfortable owning a critical subsystem end to end.

Responsibilities

• Build and evolve the agent harness and orchestration that turns an LLM into a reliable autonomous pentester.
• Design the tools and tool-shaped feedback the agent uses to probe and exploit, with schema-constrained outputs.
• Translate offensive expertise into repeatable agent capabilities in collaboration with attackers.
• Own and grow the evaluation infrastructure and regression detection to monitor improvement.
• Manage LLM inference in production: model selection, prompts, and cost/latency control (AWS Bedrock).
• Ensure production safety and no false-positives, with verifiable findings.

Skills

Python

LLM-powered apps

Agent reliability

Evaluation harness

Prompt engineering

Testing & observability

Ownership mindset

Job description

Horizon3.ai is building an autonomous web application penetration tester that operates production-safe, autonomous attacks and scales across complex environments. We seek an engineer who can convert deep offensive expertise into reliable agent capabilities, focusing on orchestration, tooling, and verification to avoid false positives.

You’ll own the agent harness, evaluation infrastructure, and production LLM management, while ensuring deterministic behavior and scalable performance in a