Security Engineer

Mercor•San Francisco, CA

4d•Onsite

About The Position

You’ll work on security across a platform that processes large volumes of sensitive data, including AI training pipelines, expert payments, and enterprise integrations. This is a hands-on engineering role focused on building and improving the systems that keep Mercor secure. We use AI heavily in our own security work. You should be comfortable building alongside AI code-gen tools, using LLMs to accelerate threat analysis, and automating away the repetitive work that slows security teams down. If you're the kind of engineer who writes a script instead of filing a ticket, you'll fit in here. We're in-person five days a week at our San Francisco headquarters, with first Fridays remote. New York could be an option for the right candidate.

Requirements

You've built security tooling or automation in a previous role - not just operated existing tools
Strong in Python, Go, or TypeScript - you ship code, not slide decks
Experience hardening cloud infrastructure (AWS preferred) - VPCs, IAM, container security
You understand application security at the code level - can review a PR for auth bugs, not just run a scanner
Comfortable with detection engineering - writing rules, tuning alerts, reducing noise
You've done incident response and know what it means to be on-call when things break
5+ years of professional experience in security engineering, software engineering, or a related builder role

Nice To Haves

Experience at a high-growth startup or fast-moving engineering org
Familiarity with AI/ML security - model access controls, training data protection, prompt injection
Offensive security skills - pen testing, bug bounty, red team experience
Contributions to open source security tools
You've built something from scratch that a team still uses

Responsibilities

Detection and response pipelines that catch real threats, not checkbox alerts
Security automation that replaces manual processes - if you're doing something twice, automate it
Infrastructure hardening across AWS, Kubernetes, and our production environment
Identity and access controls for a platform serving 300K+ experts and enterprise clients
Application security tooling integrated into CI/CD - shifting security left without slowing down deploys
Incident response runbooks and tooling - when something breaks, you'll own the fix end-to-end