AI Platform Engineer, Backend (Agentic Engineering)

5 months ago

H1B sponsor

Responsibilities

Own the foundations for company-wide LLM usage, including cost visibility and controls, data privacy, identity and access, provider routing, and security.
Design sandboxing, orchestration, audit, and guardrail layers for product agents.
Build defenses and controls for prompt injection, scoped credentials, kill switches, multi-tenant isolation, and runaway costs.
Design resource, credential, token, scheduling, quota, fairness, and observability models for agent workloads.
Build coding-agent infrastructure that can review and ship code, automate CI, refactor codebases, and run as background workers.
Enable internal and deployed teams to run agents safely against existing tools with appropriate credentials, scheduling, memory, and audit capabilities.
Own architecture, implementation, rollout, observability, on-call, and iteration based on user feedback.
Partner with security, infrastructure, and product teams to make agent deployments safe by default.

5+ years of production backend systems experience.
Deep proficiency in at least one of Python, TypeScript, Go, or Rust.
Strong distributed-systems fundamentals, including consistency, idempotency, retries, failure modes, queueing, and scheduling.
Experience designing and operating APIs and services used by other engineers.
Track record building shared infrastructure, internal platforms, or developer-facing services adopted by real users.
Strong developer-experience and maintainability judgment, including abstraction-boundary design.
Ability to own the full lifecycle from design and MVP delivery through hardening and company-wide adoption.
Experience operating services with uptime and operational responsibility, including observability, incident response, and SLOs.
Cloud-native experience with Kubernetes, infrastructure-as-code, OAuth/OIDC, and secrets management.
Preferred experience with LLM infrastructure, agent frameworks, tool-use systems, sandboxed code execution, prompt-injection defenses, multi-tenant or regulated deployments, or open-source AI infrastructure and developer platforms.

11-50 employees

We're building an AI platform and applications for the world's most important institutions. Learn more at https://brain.co/