Staff+ Software Engineer, Backend and Infra

The role

You will design and build scalable infrastructure to orchestrate LLM calls and build an AI reliability platform. You will collaborate with research teams and interface with customers to address pain points and design workflows.

What you will do

• Design and build scalable infrastructure and systems to orchestrate tens of thousands of LLM calls per second
• Collaborate closely with the research team and deploy innovative new models
• Interface directly with customers to understand and address their pain points
• Design intuitive new user workflows
• Guide customers on the evaluation methodology Haize has pioneered
• Act as a technical leader across the organization
• Mentor top talent
• Set technical direction and vision for the company
• Develop and uphold processes and best practices

Required qualifications

• 7+ years of industry experience
• 3+ years leading large initiatives as an engineer or manager
• Proven track record of designing, building, and scaling complex distributed systems
• Strong proficiency in one or more backend programming languages (e.g. Python, Go, Rust)
• Deeply knowledgeable about modern cloud infrastructure, including technologies like Kubernetes and Terraform, and platforms like AWS, GCP, and Azure
• Strong understanding of relational and NoSQL databases (Postgres, MongoDB, and Redis)
• Pragmatic, outcomes-focused approach to solving ambiguous technical and business problems
• Bachelor’s degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience

Nice to have

• Experience doing applied AI/ML
• Strong familiarity with security and privacy best practices
• Built and/or operated large scale, GPU-based model training and serving infrastructure
• Experience working on multi-datacenter, cloud-agnostic systems

Benefits

• equity
• benefits

Before you apply

• If you are exceptional, we will sponsor US visa

About the company

Haize Labs takes AI-based applications from proof-of-concept to production by rigorously, proactively, and continuously fuzz-testing them. Since 2024, they have amassed customers including OpenAI, Anthropic, and AI21.