> Markdown version of [/jobs/ext/4608315-security-engineer](https://www.wearedevelopers.com/jobs/ext/4608315-security-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Security Engineer

- **Company:** Whatnot Inc.
- **Location:** San Francisco, CA, United States (Remote available)
- **Experience:** Experienced
- **Contract:** Permanent contract
- **Skills:** Active Directory, Application Programming Interfaces (APIs), Artificial Intelligence, Software System Penetration Testing, Cloud Computing Security, Code Review, Computer Programming, Continuous Integration, Information Leak Prevention, Identity and Access Management, Intrusion Detection and Prevention, Python (Programming Language), Regression Testing, Red Team (Cyber Security), Data Logging, Large Language Models, Mitre Att&ck, Kubernetes, Bug Reporting, Purple Team (Cyber Security), Blue Team (Cyber Security)
- **Published:** November 6, 2026
- **Apply:** https://startup.jobs/security-engineer-red-team-whatnot-com-10694403

## About the Role

Curious about who thrives at Whatnot? We've found that low ego, a growth mindset, and leaning into action and high impact goes a long way here.

As our next Red Team Engineer you'll find weaknesses in our infrastructure before real attackers do. You will use AI tools and agents to scale offensive testing, build the harnesses that run it, and turn findings into fixes across cloud, network, application, identity and CI/CD layers.

 * 3+ years in offensive security, red teaming, or penetration testing, with real engagements against production-scale environments.
 * Strong programming skills in Python; you write your own tooling instead of only running others.
 * Hands-on experience building LLM-powered workflows or agents (tool use, orchestration, evaluation, guardrails), not just prompting a chat window.
 * Deep knowledge of at least two areas: cloud security and IAM abuse, Active Directory or identity attacks, Kubernetes and container escapes, web and API exploitation, CI/CD and supply chain attacks.
 * Working knowledge of AI-specific risks: prompt injection, insecure tool use, data leakage, excessive agent permissions, and model supply chain issues.
 * Sound judgment on scope, safety, and blast radius; you know how to test production-adjacent systems without causing outages.
 * Clear written communication: reports an engineer can act on and an executive can skim.

## Description

* Plan and run authorized red team operations across cloud (AWS), Kubernetes, networks, identity systems, SaaS, and internal applications, under written rules of engagement.
 * Build and maintain AI-assisted testing harnesses: agent loops that run recon, attack-path discovery, fuzzing, and exploit validation with logging, scoping controls, and human approval gates.
 * Use LLMs and agentic tools to speed up vulnerability research, payload and test-case generation, code review, and report drafting, and measure where they help and where they mislead.
 * Red team our own AI systems: prompt injection, tool and connector abuse, data exfiltration, model and agent permission boundaries, and supply chain risk in AI dependencies.
 * Emulate realistic adversaries using MITRE ATT&CK and ATLAS, chaining findings into end-to-end attack paths rather than isolated bugs.
 * Run purple team exercises with detection engineering and incident response, and tune detections based on what your tests evade.
 * Write clear findings with reproduction steps, impact, and recommended fixes; track remediation to closure and retest.
 * Automate regression tests so fixed weaknesses stay fixed, and feed results into CI/CD pipelines.
 * Document and share techniques, tooling, and lessons learned with engineering teams.
 * Work collaboratively with Whatnot's Blue Team on Purple Team engagements to help strengthen Whatnot's overall security posture.