Senior Software Engineer — DevOps & Platform, AI Systems

Who You Are 

A software engineer who chose infrastructure. You write real code — not just glue scripts. You'd be dangerous on a product team, but you'd rather build the systems that make every product team faster. Platforms, not tickets. 

An agent orchestrator, not a typist. You don't want to be a faster keyboard. You manage agents to do the heavy lifting — generating IaC, triaging alerts, writing the first draft of the runbook — while you keep your hands on the architecture and the judgment calls. The terminal is a cockpit for orchestration. 

Productively lazy. Your instinct is to eliminate toil at the root. If a human is clicking through a console or babysitting a deploy, you replace them with code or an agent. Your biggest bottleneck should be deciding what to automate next — while the automation runs itself. 

Fundamentals first. Linux, networking, distributed systems, how things actually fail under load. You understand the machine, not just the tool that wraps it. When the pipeline breaks at 2am, you can read the trace and fix the cause — and when an agent proposes a fix, you can tell whether it's right. 

First-principles thinker with vision. You design for the failure mode, not the happy path. You question the assumptions, simplify the system, and build for the next order of magnitude — not just today's load. 

High agency. You don't wait for perfect specs or permission. You find a path, propose it, and move. Large organizations have walls; you figure out which ones to go through, around, or remove — and you do it constructively. 

What You'll Do 

• Build the platform. Design and operate the cloud-native foundation on GCP — compute, networking, observability, and the golden paths other teams build on. 

• Automate everything that shouldn't be manual. Infrastructure-as-code, self-service pipelines, auto-remediation. If a human repeats it, you write the code — or point an agent at it — to stop it. 

• Orchestrate agents for ops. Use AI to draft IaC, generate tests, triage incidents, and summarize signal from noise. You direct, review, and harden — agents do the grunt work. 

• Own CI/CD end to end. Fast, reliable, reproducible delivery. Build the pipelines that let engineers ship multiple times a day with confidence. 

• Engineer reliability in. SLOs, monitoring, alerting that means something, and incident response that gets calmer every quarter. Make on-call boring. 

• Secure the path. Bake security and compliance into the platform so doing the right thing is the easy thing — secrets management, least privilege, supply-chain hygiene. 

• Support AI workloads. Build the infrastructure that runs LLM and agentic systems in production — scaling, cost control, and the deployment patterns AI apps actually need. 

What You Bring 

• A real programming language, deep — Python and/or Go preferred. You write production code, not just YAML. 

• GCP — Cloud Run, GKE, Cloud Functions, networking, IAM. Designed, deployed, and operated at scale. 

• IaC & CI/CD — Terraform, Cloud Build, GitHub Actions, or equivalent. Reproducible, version-controlled, peer-reviewed infrastructure. 

• Containers & orchestration — Docker and Kubernetes in production, including the parts that break. 

• Observability — metrics, logs, tracing. You can find the needle, not just collect haystacks. 

• AI-assisted development — fluent with modern AI coding tools and agent workflows, and clear-eyed about where they help and where they don't. 

• AI workload experience — bonus for running LLM/agentic systems, GPUs, or model-serving infrastructure in production. 

• Reliability & security fundamentals — SLOs, incident response, and security-by-default in pipelines.