Senior Security Engineer, Platforms & AI - Remote - Latin America

Azumo builds and operates production AI systems for companies ranging from seed-stage startups to Meta. We are hiring a Senior Security Engineer to own the technical security posture of those systems once they are live: the infrastructure they run on, the pipelines that feed them, and the agents that act on their output. The role is fully remote across Latin America, aligned to your client's working day.

You will not be writing policy. Azumo has shipped production AI since 2016, and the work here is in the systems themselves — hardening infrastructure, closing vulnerabilities, and building the guardrails that keep AI-driven workflows safe in front of real users.

Where this role sits

Most engineering organizations split ownership by what's built: pipelines and data infrastructure, the model or the method behind a decision, the product itself, and — once AI systems are live — what they actually do in front of users. Security isn't one more slice next to those . It's the discipline that has to hold inside every one of them, and on any team large enough to carry a dedicated specialist, this is the role that owns it .

One question places the boundary: when a system is breached, was a control never built, or did someone get past one that existed? The first is an engineering gap, and wherever it sits — infrastructure, pipeline, agent, integration — it's this role's problem to close before someone outside the team finds it. The second is an incident, and this role runs it : containment, root cause, and the fix that keeps it from happening twice.

On engagements built around an embedded delivery model, this role's accountability is distinct from whoever owns delivery of the system itself: they answer for whether it works, this role answers for whether it's safe to run . Same principle either way — no handoff , one shared understanding of what "secure" means, agreed before the work starts rather than reported after.

What you will build

Platform and infrastructure hardening. Cloud infrastructure, APIs, internal tooling, and CI/CD pipelines: encryption, secrets management, logging, and access controls that are built in, not bolted on, and stay correct as the systems around them change.

Vulnerability management and offensive security. Scanning, penetration testing, adversarial testing, threat modeling, and manual review across everything Azumo ships into a client's environment — with remediation you drive to closed, not to ticket.

AI and agent security. Prompt-injection defense, adversarial-input testing, and guardrails for the tool-calling and agentic systems the AI Engineer lane builds — the same production behavior, hardened against misuse rather than measured for accuracy. Bounded, observable, auditable, and safe to fail.

The standing audit. Azumo runs an automated security, cost, and architecture audit across every client codebase, graded by severity down to the file and line, on day one and every day after. You own it — what it checks, how it's graded, how often it's wrong.

Monitoring, detection, and incident response. Alerting and detection for what you harden, investigation when something looks wrong, and root cause when it turns out to be something.

Work inside the client's environment. Their repositories, their tooling, sometimes their compliance review. Azumo is SOC 2 certified, client code stays in client repositories, and some engagements carry additional requirements such as HIPAA.

How we work

Our engineers build with AI every day . Claude Code, Codex, and similar tools are part of the standard toolchain here, not an experiment. The audit you'll own runs across the whole codebase on day one and every day after, grading security, cost, and architecture findings by severity with the exact file and line, so a small team can move quickly without quality drifting. We stay vendor-neutral across OpenAI, Anthropic, and open-weight models, and we run Valkyrie, our own production layer, when a single interface to any model