Remote Full-Stack Engineer III — AI Gateway

Ngrokinc; San Francisco (CA); Hybrid; USD 186,000 - 248,000; Full time

Benefits offered by this job

Health insurance

Retirement matching

Flexible time off

Parental leave

Professional development budget

Home office stipend

Co-working stipend

Lunch on us

Company offsites

Regular feedback

Job summary

ngrok Inc. is a cloud networking platform powering AI/ML traffic gateways and secure API delivery. We build systems to identify, control, and understand AI traffic in real time, with scalable policies and usage metrics.

You’ll join the AI Gateway Team to ship reliable, observable traffic control in production. You’ll work on a fast, type-safe stack (Go/TypeScript) across AWS, Kubernetes, and modern tooling, collaborating with product and design to create safe, abstractions that minimize provider

Qualifications

• Experience with statically typed languages and modern cloud platforms.
• Experience with AI/LLMs and edge-case handling.

Responsibilities

• Build and evolve the AI Gateway components to classify AI traffic in real time.
• Own AI traffic policy enforcement with rate limits and access rules.
• Track AI usage, tokens, and metadata for customers.
• Make AI behavior observable and explainable with clear signals.
• Design abstractions to hide infrastructure complexity for users.
• Ship production-grade systems with cross-team collaboration.

Skills

Go

Rust

C++

Java

TypeScript

React

HTML

CSS

AI/LLMs

Dev experience

System design

Tools

AWS

Kubernetes

PostgreSQL

Kafka

Protobuf

Terraform

Helm

Buildkite

GitHub

Job description

ngrok Inc. is a cloud networking platform powering AI/ML traffic gateways and secure API delivery. We build systems to identify, control, and understand AI traffic in real time, with scalable policies and usage metrics.

You’ll join the AI Gateway Team to ship reliable, observable traffic control in production. You’ll work on a fast, type-safe stack (Go/TypeScript) across AWS, Kubernetes, and modern tooling, collaborating with product and design to create safe, abstractions that minimize provider