Lead AI Agent Security Architect

Neura Market; Foster City (CA); Hybrid; USD 170,000 - 250,000; Full time

Benefits offered by this job

Competitive Salary & Equity

401(k) with match

Health, Dental, Vision

Paid Parental Leave

Flexible Time Off

In-Office Amenities

Job summary

Replit seeks an AI Agent Security Architect to be the principal technical authority for securing autonomous agent workflows. You will design runtime defenses, guardrails, and sandboxing architectures that govern agents executing code, invoking tools, and reasoning across our platform.

You will lead high-impact AI security initiatives, bridge non-deterministic AI behavior with cybersecurity controls, and collaborate with engineering and executive leadership on risk management and compliance.

Qualifications

• 6+ years in security engineering or security architecture with AI/ML security focus.
• Deep understanding of agentic architectures and protocols (MCP, LangChain, ReAct).
• Experience with AI security threats: prompt injection, hijacking, data poisoning.
• Proficiency in Python, Go, or TypeScript/JavaScript with security tooling.
• Ability to document security architecture and communicate to engineers and executives.
• Experience contributing to an enterprise Cybersecurity Risk Register.

Responsibilities

• Define AI agent security strategy and long-term execution plan.
• Design guardrails, semantic firewalls, and real-time verification filters.
• Create sandboxing for dynamic code execution (microVMs, WebAssembly, containers).
• Lead threat modeling and red-teaming for non-deterministic systems.
• Develop least-privilege and HITL authorization models.
• Support risk management and maintain the Cybersecurity Risk Register.
• Establish auditability and telemetry standards for incident response.

Skills

AI security

Security architecture

Threat modeling

Python/Go/TS

Security tooling & guardrails

Executive communication

Education

B.S. in CS or related field

Tools

LangChain

Model Context Protocol (MCP)

ReAct

Vector databases

WebAssembly

Job description

Replit seeks an AI Agent Security Architect to be the principal technical authority for securing autonomous agent workflows. You will design runtime defenses, guardrails, and sandboxing architectures that govern agents executing code, invoking tools, and reasoning across our platform.

You will lead high-impact AI security initiatives, bridge non-deterministic AI behavior with cybersecurity controls, and collaborate with engineering and executive leadership on risk management and compliance.