Member of Technical Staff, Model Serving Infrastructure

Position Overview

• Location Type: Remote
• Employment Type: Full-Time
• Salary: Not provided

Cohere is dedicated to scaling intelligence to serve humanity by training and deploying frontier models for developers and enterprises. They foster a culture of hard work, rapid iteration, and customer focus. The company is seeking Members of Technical Staff to join the Model Serving team. This team is responsible for the development, deployment, and operation of the AI platform that delivers Cohere’s large language models via API endpoints. The role involves collaborating with various teams to deploy optimized NLP models into production, ensuring low latency, high throughput, and high availability. Additionally, the position includes direct customer interaction and the creation of customized deployments.

Requirements

• Experience: Minimum of 5 years of engineering experience managing production infrastructure at a large scale.
• System Design: Proven experience in designing large, highly available distributed systems, with a focus on Kubernetes and GPU workloads.
• Cloud Platforms: Hands-on experience with Kubernetes development and production support across platforms such as GCP, Azure, AWS, OCI, multi-cloud environments, and on-prem/hybrid setups.
• Linux Expertise: Demonstrated experience in designing, deploying, supporting, and troubleshooting complex Linux-based computing environments.
• Resource Management: Experience in managing compute, storage, and network resources, including cost optimization.
• Collaboration & Troubleshooting: Excellent skills in collaboration and troubleshooting, essential for building mission-critical systems and ensuring smooth operations.
• Adaptability: Demonstrated grit and adaptability to effectively solve complex technical challenges.
• Accelerator Knowledge: Familiarity with the computational characteristics of accelerators (GPUs, TPUs, and/or custom accelerators) and their impact on inference latency and throughput.
• Distributed Systems Understanding: A strong understanding or practical experience with distributed systems.
• Programming Languages: Proficiency in Golang, C++, or other languages suitable for high-performance, scalable server development.

Responsibilities

• Develop, deploy, and maintain the AI platform that serves Cohere’s large language models.
• Collaborate with cross-functional teams to deploy optimized NLP models into production environments.
• Ensure the low latency, high throughput, and high availability of deployed AI models.
• Engage directly with customers and develop customized deployment solutions to meet their specific requirements.
• Manage and optimize compute, storage, and network resources and associated costs.
• Troubleshoot and resolve complex technical issues to maintain system stability and performance.
• Contribute to the enhancement of Cohere’s models and the value they deliver to customers.

Application Instructions

• Not provided in the description.

Company Information

• Mission: To scale intelligence to serve humanity.
• Focus: Training and deploying frontier models for developers and enterprises.
• Culture: Emphasizes building, responsibility for increasing model capabilities, working hard and moving fast, and valuing diverse perspectives.