AI Systems / Agent Runtime Engineer (Intern / New Grad / Full-time)

AI Systems / Agent Runtime Engineer (Intern / New Grad / Full-time)

AI Systems / Agent Runtime Engineer (Intern / New Grad / Full-time)

AI Systems / Agent Runtime Engineer (Intern / New Grad / Full-time)

Location: Singapore / Remote
Company: Polygala



About Polygala

We're building the cloud for AI agents. Today's cloud was built for human developers — shaped around consoles, dashboards, and week-long provisioning cycles that assume a person in the loop. We're building the cloud for AI agents: millisecond provisioning, autonomous scaling, and a programmatic surface agents can drive end to end. Polygala is rethinking infrastructure from the ground up for an AI-native world. Our stack includes components like BoxLite — an embeddable runtime for secure agent execution: https://github.com/boxlite-ai/boxlite



What You’ll Do

Runtime & Sandbox (Core)

Build and improve sandboxed execution for AI agents, with a focus on reliability, performance, and security. Work on core problems such as isolation, resource control, persistence, snapshotting, and multi-agent execution.


Systems & Infrastructure

Contribute to runtime internals and the broader infrastructure layer that supports agent workloads. You’ll work across areas like containers, OCI, virtualization concepts, and system efficiency, helping improve overall stability and operational quality.


Developer Experience

Shape the way developers interact with the stack through better CLI tooling, SDKs, APIs, and documentation. You’ll help make powerful infrastructure feel simple, understandable, and easy to integrate into real agent products.


Open Source & Ecosystem

Contribute directly to BoxLite and help grow its developer ecosystem. That includes improving the open-source project itself, engaging with builders, and supporting integrations, examples, and community adoption.



Requirements

You have a strong computer science or engineering background and are comfortable writing production-quality code. We’re especially interested in people with solid skills in Rust, Go, Python, or C++, and who can think clearly about backend systems, infrastructure, and low-level technical tradeoffs.


Nice to Have

Experience with Docker, OCI, or container-based workflows is a strong plus. Familiarity with operating systems, virtualization, security boundaries, open-source projects, or infrastructure and agent system design would also be valuable in this role.


What We Care About

We care most about what you’ve actually built, how well you understand systems, and how quickly you learn. Strong resumes are welcome, but demonstrated ability, technical depth, and builder instinct matter much more than credentials alone.


Why Join

You’ll work on core infrastructure for the next generation of AI systems, not just application-layer tooling. This role offers deep exposure to systems problems, meaningful open-source impact, high ownership from day one, and a path toward full-time conversion for the right person.


How to Apply

Send your GitHub, selected projects, and anything else that best shows how you build.


Dorian
dorian@boxlite.ai


Strong builders hear back fast.


Bonus (Optional)

As an optional extra, you can design a safe execution environment for AI agents and share how you would think about isolation, permissions, resource limits, and developer usability.

Location: Singapore / Remote
Company: Polygala



About Polygala

We're building the cloud for AI agents. Today's cloud was built for human developers — shaped around consoles, dashboards, and week-long provisioning cycles that assume a person in the loop. We're building the cloud for AI agents: millisecond provisioning, autonomous scaling, and a programmatic surface agents can drive end to end. Polygala is rethinking infrastructure from the ground up for an AI-native world. Our stack includes components like BoxLite — an embeddable runtime for secure agent execution: https://github.com/boxlite-ai/boxlite



What You’ll Do

Runtime & Sandbox (Core)

Build and improve sandboxed execution for AI agents, with a focus on reliability, performance, and security. Work on core problems such as isolation, resource control, persistence, snapshotting, and multi-agent execution.


Systems & Infrastructure

Contribute to runtime internals and the broader infrastructure layer that supports agent workloads. You’ll work across areas like containers, OCI, virtualization concepts, and system efficiency, helping improve overall stability and operational quality.


Developer Experience

Shape the way developers interact with the stack through better CLI tooling, SDKs, APIs, and documentation. You’ll help make powerful infrastructure feel simple, understandable, and easy to integrate into real agent products.


Open Source & Ecosystem

Contribute directly to BoxLite and help grow its developer ecosystem. That includes improving the open-source project itself, engaging with builders, and supporting integrations, examples, and community adoption.



Requirements

You have a strong computer science or engineering background and are comfortable writing production-quality code. We’re especially interested in people with solid skills in Rust, Go, Python, or C++, and who can think clearly about backend systems, infrastructure, and low-level technical tradeoffs.


Nice to Have

Experience with Docker, OCI, or container-based workflows is a strong plus. Familiarity with operating systems, virtualization, security boundaries, open-source projects, or infrastructure and agent system design would also be valuable in this role.


What We Care About

We care most about what you’ve actually built, how well you understand systems, and how quickly you learn. Strong resumes are welcome, but demonstrated ability, technical depth, and builder instinct matter much more than credentials alone.


Why Join

You’ll work on core infrastructure for the next generation of AI systems, not just application-layer tooling. This role offers deep exposure to systems problems, meaningful open-source impact, high ownership from day one, and a path toward full-time conversion for the right person.


How to Apply

Send your GitHub, selected projects, and anything else that best shows how you build.


Dorian
dorian@boxlite.ai


Strong builders hear back fast.


Bonus (Optional)

As an optional extra, you can design a safe execution environment for AI agents and share how you would think about isolation, permissions, resource limits, and developer usability.