AI Agent Security Architect
Estimated pay Replit didn’t post pay for this job. This is our estimate: the median of 90 posted ranges for similar jobs at any level in the same country. Ask the employer for the actual range.
The posting
Overview
Replit is the agentic software creation platform that enables anyone to build applications using natural language. With millions of users worldwide, Replit is democratizing software development by removing traditional barriers to application creation.
About the job
We are looking for an AI Agent Security Architect to function as the primary technical authority for Replit’s autonomous and AI agent security blueprint. In this critical role, you will design, implement, and maintain the runtime defense systems, guardrail frameworks, and sandboxing architectures that govern AI agents executing code, invoking tools, and reasoning across our platform. You will be a key technical contributor—leading high-impact AI security initiatives and bridging the gap between non-deterministic AI behavior and rigorous cybersecurity controls for both engineering and executive leadership.
What you’ll do
- AI Agent Security Strategy & Technical Execution
- AI Agent Security Blueprint: Define the long-term vision and architectural patterns for securing autonomous agent workflows, Model Context Protocol (MCP) integrations, multi-turn reasoning loops, and multi-agent coordination.
- Runtime Guardrails & Policy Enforcement: Architect and deploy dynamic input/output guardrail systems, semantic firewalls, and real-time intent verification filters to prevent goal hijacking, system prompt leaks, and indirect prompt injections.
- Agent Execution & Tool Sandboxing: Partner with Infrastructure and AppSec teams to design secure, short-lived, micro-isolated environments (e.g., microVMs, WebAssembly, container sandboxes) where agents can dynamically execute code, run shell commands, and interact with host operating systems safely.
- Agentic Threat Modeling & Red Teaming: Conduct specialized threat modeling against non-deterministic systems. Lead automated and manual AI red-teaming initiatives to uncover vulnerabilities in RAG context pipelines, vector stores, and tool-calling interfaces.
- Identity, Delegation & Least Privilege: Architect fine-grained permission models and step-up Human-in-the-Loop (HITL) authorization patterns for agents acting on behalf of users—ensuring agents never exceed the delegated privileges of the end user or misuse API keys.
- Context & Memory Boundary Security: Design strict data isolation and poisoning prevention controls for vector databases, retrieval-augmented generation (RAG) pipelines, and long-term conversation memories across multi-tenant workloads.
- Risk Management & Cross-Functional Enablement
- Maintain the AI Security Source of Truth: Define, document, and maintain authoritative secure design patterns and SDKs for engineering teams building internal or customer-facing AI agent capabilities.
- Contribution to Risk Register: Identify, quantify, and document non-deterministic and agentic security risks (e.g., OWASP Top 10 for LLMs & AI Agents, MITRE ATLAS), ensuring they are accurately represented in the Cybersecurity Risk Register.
- Auditability & Observability: Design tamper-evident logging and telemetry standards for agent reasoning chains, tool execution history, and context assembly to support incident response, forensics, and GRC requirements.
- Compliance & Sales Enablement: Partner with GRC to translate complex AI security controls into enterprise-ready audit documentation (e.g., ISO 42001, NIST AI RMF, EU AI Act readiness) and support Sales on complex enterprise security inquiries regarding AI safety.
Requirements
- 6+ years of experience in security engineering or security architecture, with at least 2+ years dedicated specifically to AI/ML security, LLM application security, or securing agentic frameworks.
- Deep understanding of agentic architectures and protocols (e.g., Model Context. Protocol (MCP), LangChain, AutoGen, CrewAI, ReAct frameworks, and vector database technologies).
- Hands-on expertise with threat vectors unique to agentic AI: Indirect Prompt
- Injection, Goal Hijacking, Tool Parameter Tampering, Data Poisoning, and Context Contamination.
- Strong background in applying safety standards and risk frameworks such as OWASP Top 10 for LLMs & AI Agents, NIST AI RMF, and MITRE ATLAS.
- Proficiency in Python, Go, or TypeScript/JavaScript with a proven track record of reviewing code and building functional security tooling or guardrails.
- Experience authoring, maintaining, and evangelizing technical security architecture documentation.
- Exceptional ability to communicate complex non-deterministic AI risks to both deep technical engineers and executive stakeholders.
- Proven track record of contributing to an enterprise Cybersecurity Risk Register.
Nice to have
- Experience designing runtime sandboxing and isolation technologies (e.g., Firecracker, gVisor, Docker, WebAssembly) for dynamic code execution environments.
Pay and benefits
- 💰 Competitive Salary & Equity
- 💹 401(k) Program with a 4% match ( US Only )
- ⚕️ Health, Dental, Vision and Life Insurance
- 🩼 Short Term and Long Term Disability
- 🚼 Paid Parental, Medical, Caregiver Leave
- 🏝 Flexible Time Off (FTO) + Holidays
- 🚗 Commuter Benefits ( In-Office & US Only )
- 📱 Monthly Wellness Stipend
- 🧑💻 Autonomous Work Environment
- 🖥 In Office Set-Up Reimbursement ( In-Office Only )
- 🚀 Quarterly Team Gatherings
- ☕ In Office Amenities ( In-Office Only )
About the company
- Self-driving Company
- Replit Agent at Scale
- AI Adoption
- Build Open-Source Apps
Interviewing + Culture at Replit
- Operating Principles
- Reasons not to work at Replit
To achieve our mission of making programming more accessible around the world, we need our team to be representative of the world. We welcome your unique perspective and experiences in shaping this product. We encourage people from all kinds of backgrounds to apply, including and especially candidates from underrepresented and non-traditional backgrounds.
Read from the employer’s careers page. Tags are matched from the posting’s text, so check the employer’s page for the full requirements.
Check your fit
Paste your resume to see how it lines up with this posting, what could screen you out and what to prepare.
Your resume never leaves this browser. The check runs in your browser. Nothing you paste is sent to us or anyone else.
- Pay is our estimate, not posted
Paste your resume above and press Check my fit to see your score.
Years of experience
The posting asks for 6+ years. Paste your resume to compare.
Resume layout
A simple layout reads best: plain headings, no tables or columns, and no text inside images or headers.
Application security
Walk through a vulnerability you found or fixed in code: the root cause, the fix, and how you kept it from coming back.
From the posting: Application security, OWASP, Threat modeling
GRC and compliance
Take one control from a framework you know and explain how you test it and what evidence shows it works.
From the posting: GRC, Risk assessment, NIST frameworks, RMF and ATO
AI security
Explain how you would test an AI feature for prompt injection and data leaks, and what controls you would put around it.
From the posting: AI security
Incident response
Walk through one incident you handled end to end: how it was found, how you contained it, how you recovered and what changed after.
From the posting: Incident response, Digital forensics
Threat intelligence
Explain how you turn a threat report into something the team acts on, such as a detection, a block or a hunt.
From the posting: MITRE ATT&CK
Offensive security
Walk through one engagement: scope, how you got your first foothold, what you found and how you wrote it up for the client.
From the posting: Red teaming
Your experience
The posting asks for 6+ years. Have a two-minute walk through your background that ends with why this job is the next step.
The score compares words and numbers, so treat it as a guide, not a verdict. Requirements are read from the posting’s text; check the employer’s page for the full list.