Software Engineer II - Slack Developer Workspaces

Momentum; San Francisco (CA); On-site; USD 140,000 - 210,000; Full time

Job summary

Slack is building fast, isolated sandboxes for AI agents to run a full dev toolchain. This role owns production systems early, handling pager duties and guiding safe agent execution at scale.

You will scale the sandbox fleet, harden isolation, and improve remote dev environments. You should have strong Go/Python, AWS and Terraform skills, and be comfortable with on-call responsibilities.

Qualifications

• 3+ years of experience building and operating production systems.
• Experience in infrastructure, SRE, DevOps or platform engineering, with on-call responsibility.
• Strong AWS knowledge: EC2, networking, IAM.
• Proficient in Go or Python; infrastructure as code with Terraform.
• Experience with AI coding tools and integrating them into workflows.
• Ownership mindset; proactive in fixing issues.

Responsibilities

• Scale AI sandbox fleet with capacity planning, safe rollouts and image builds.
• Harden isolation between agents and Slack systems: network, secrets handling and patching.
• Run and improve remote dev environments, including Terraform and VM images.
• Build observability: SLOs, dashboards, and logs.
• Join weekly on-call rotation; fix issues and automate the fix.
• Work with AI agents; design platform needs they rely on.

Skills

Production experience

Go/Python

AWS

Terraform

On-call experience

AI tooling

Tools

Terraform

KVM/Firecracker

Job description

The Experience

Developer Workspaces runs the computers Slack engineers write code on - and more and more, the ones AI agents write code on. Every week, hundreds of engineers build and test Slack in the remote development environments we run on AWS and Kubernetes. Now we're building the next layer: fast, isolated sandboxes where coding agents can run a dev toolchain, open pull requests, and shut down when they're done.

If most of Slack's code is going to be written by agents, it'll be written on infrastructure this team owns.

We're a small team with a lot of surface area. We're strong on backend engineering, and we need more depth in infrastructure and operations. That's this role. You'll own production systems early, carry the pager, and help decide how agents run safely at Slack.

What You'll Actually Be Doing

• Scale our AI sandbox fleet into something the whole company depends on - capacity planning, safe rollouts, image builds, and the alerts that tell us when it's breaking.
• Harden the isolation between agents, sandboxes and the rest of Slack: network separation, egress controls, secrets handling, and patching.
• Run and improve the remote dev environments our engineers use every day, from Terraform and machine images to the automation that pauses idle environments so they don't burn money.
• Build observability we can make decisions with - SLOs, dashboards, and logs that land where we can find them.
• Join our weekly on-call rotation once you've ramped up. Most pages are low urgency. When something does break, we fix it and then automate the fix.
• Use AI agents heavily in your own work. You're building the platform they run on, so you should know what they need.

You're Our Person If...

• You have 3+ years of experience building and operating production systems.
• You come from infrastructure, SRE, DevOps or platform engineering, and you've been on call for something that mattered.
• You know AWS well - EC2, networking, IAM - and you manage infrastructure as code with Terraform or something like it.
• You write solid code in Go or Python. Everything we do ships as code, and your infrastructure instincts shape how you write it.
• You use AI coding tools every day and have opinions about where they fall short.
• You take ownership. When something breaks, you don't wait for someone to assign it to you.

Even Better If...

• You've worked with Firecracker, KVM, or other virtualization.
• You've run Kubernetes in production.
• You have a security background: VM or container isolation, vulnerability management, threat modeling.
• You're comfortable in Linux internals and networking.
• You've worked on a developer productivity or platform team.