Senior Software Engineer

LightSprint; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

LightSprint is seeking a senior software engineer in San Francisco to own the sandboxing infrastructure that launches and retires agent-driven environments. You will move between application code and cloud systems to make sandboxes fast, observable, and safe for review.

You will build, operate, and improve the end-to-end environment lifecycle, working closely with product and engineering teams to turn complex workflows into robust, scalable software that supports customer code changes.

Qualifications

• Built and operated production infrastructure for developer tools or cloud apps.
• Hands-on with ephemeral sandboxes or remote execution.
• Understanding isolation: containers/VMs, secrets, networking.
• Ability to build APIs, job orchestration, state management.
• Ability to diagnose incidents from logs/metrics/traces.
• Clear technical planning and ownership after ship.

Responsibilities

• Build reliable sandboxes and improve provisioning and cleanup.
• Make previews dependable and diagnose setup failures.
• Design for safe execution: permissions, secrets, networking.
• Connect agent changes, repositories, checks, and reviews.
• Operate and improve instrumentation, alerts, and runbooks.
• Collaborate with product and engineering to refine workflows.

Skills

Production infrastructure

Cloud platforms

Troubleshooting

System design

Cross-functional collaboration

Security/compliance awareness

Technical writing

Tools

Containers (Docker)

VMs

Kubernetes

GitHub workflows

CI pipelines

Secrets management

Monitoring & tracing

Job description

About us

LightSprint brings product and engineering teams into one shared workflow for building software. Coding agents work on real codebases in isolated sandboxes, with previews people can inspect before changes go through engineering review and merge.

We want it to feel simple to the person asking for a change and dependable to the engineer responsible for the code. The infrastructure underneath has to make both possible.

The role

We’re looking for a senior software engineer to work on the systems that launch, run, and retire sandbox environments for agent-driven software changes. You’ll own infrastructure work from design to production operation, and work closely with the product and engineering team to make each session fast to start, useful to collaborate in, and safe to review.

This is a hands-on role for someone who can move between application code and cloud infrastructure. You should enjoy debugging a failed environment as much as designing the system that prevents the next one.

What you’ll do

• Build reliable sandboxes. Improve how environments are provisioned, configured, isolated, observed, and cleaned up as agents work on customer codebases.
• Make previews dependable. Help teams get working preview environments quickly, diagnose setup failures, and understand what happened when a run goes wrong.
• Design for safe execution. Think through permissions, secrets, network boundaries, resource limits, and lifecycle controls for code that runs in isolated environments.
• Connect the work to engineering review. Build and maintain the paths between an agent’s changes, the repository, checks, and the human review process.
• Operate what you build. Add the instrumentation, alerts, and runbooks needed to investigate failures and improve reliability and cost over time.
• Work across the product. Partner with engineers and product teammates to turn rough workflows and customer problems into clear technical decisions.

Requirements

• You’ve built and operated production infrastructure for developer tools, cloud applications, CI systems, or other products that run user or agent workloads.
• You have hands‑on experience with ephemerich sandboxes or remote execution. Experience with E2B, Blaxel, or a comparable platform is especially relevant.
• You understand isolation and its tradeoffs, including containers or VMs, access to secrets, networking, tenancy, and what can fail when untrusted code runs.
• You can build the application-side systems around infrastructure, such as APIs, job orchestration, state management, and integrations with code repositories.
• You can diagnose production incidents using logs, metrics, and traces, and turn what you learn into simpler, more reliable systems.
• You write clear technical plans, make pragmatic tradeoffs, and take responsibility for the behavior of a system after it ships.

Nice-to-have

• Experience running coding agents, browser-based development environments, or per-branch previews at scale.
• Familiarity with GitHub workflows, CI pipelines, dependency installation, and codebase onboarding across different stacks.
• Experience improving cold-start times, resource utilization, or cost for short-lived compute workloads.
• Experience building infrastructure for teams with security or compliance requirements.