Kosha Labs
Website:
kosha.or.kr
Company:
https://www.linkedin.com/company/koshalabsai
Seniority: Entry level
Industries: Robotics Engineering
Job details:
About Us:
Kosha Labs is building the data infrastructure for physical AI. We collect and process large-scale multimodal robotics datasets across vision, tactile sensing, motion capture, and robot demonstrations. We're backed by leading Silicon Valley investors and work with frontier AI and robotics labs.
The Role:
We're looking for a Cloud Infrastructure Engineer (OCI) to build and operate the infrastructure behind petabyte-scale multimodal datasets and 100+ Gbps data pipelines. You'll own high-throughput storage, networking, compute, and data movement across Cloud Infrastructure. This is a hands-on infrastructure role focused on maximizing throughput, reliability, and scalability while keeping infrastructure costs under control.
What you'll work on:
- Design infrastructure for ingesting, processing, storing, and delivering petabyte-scale robotics datasets.
- Build 100+ Gbps data transfer pipelines across cloud infrastructure.
- Optimize object storage for high-throughput parallel reads/writes and large-scale transfers.
- Design networking across VPCs, gateways, load balancers, and high-bandwidth compute.
- Build and operate GPU and CPU compute infrastructure for large-scale data processing.
- Develop parallel and multipart upload/download systems for moving massive datasets efficiently.
- Build infrastructure using Terraform and other IaC tooling.
- Implement monitoring, observability, failure recovery, and automated infrastructure validation.
- Profile and eliminate storage, network, compute, and I/O bottlenecks.
- Optimize infrastructure architecture and cloud spend as data volume scales
What we're looking for:
- Strong experience building and operating large-scale cloud infrastructure.
- Deep understanding of Linux, networking, storage systems, and distributed systems.
- Experience with OCI or AWS is strongly preferred.
- Experience moving or processing datasets at TB/PB scale.
- Strong understanding of TCP/IP, network throughput, parallelization, bandwidth utilization, and I/O performance.
- Experience with object storage, GPU/CPU compute clusters, Terraform, containers, and observability.
- Comfortable debugging performance bottlenecks across the entire stack.
- High agency. You should be comfortable owning infrastructure end to end and figuring things out without a playbook.
You'll enjoy this role if you like processing 100+ terabytes of data every day, being the backbone of a system at massive scale, and pushing infrastructure to its absolute limits. If you thrive on debugging hard systems problems and seeing performance improvements directly move production, you'll love this work.
Location: Bengaluru
Click on Apply to know more.
