Software Engineer, Platform & Inference

About the role

Platform engineers at Chai Discovery build the infrastructure that makes cutting‑edge molecular models fast, cheap, and reliable at scale. You will own the serving stack that transforms frontier models into a product used by scientists, ensuring low latency, high throughput, and efficient GPU utilization across a multi‑cloud fleet.

Key responsibilities

• Design, implement, and maintain the model serving platform, handling latency, throughput, batching, and autoscaling across a large GPU fleet.
• Develop tooling for turning raw models into production‑ready pipelines and provide experiment and observability support for researchers.
• Collaborate closely with model researchers, product engineers, and commercial teams to deliver reliable AI services to major pharma customers.

Required profile

• 4+ years building production systems with deep expertise in performance, distributed systems, or ML serving.
• Hands‑on experience optimizing model inference, including GPU utilization, batching, quantization, caching, or kernel‑level work.
• Platform mindset: enjoys building abstractions that accelerate other engineers and researchers.
• End‑to‑end ownership of 24/7 services, including observability, alerting, and incident response.
• Track record of both 0‑to‑1 builds and scaling to production, treating cost and efficiency as primary constraints.

Required skills

• GPU utilization
• Batching
• Quantization
• Caching
• Kernel‑level optimization
• Distributed systems
• ML serving
• Observability
• Alerting
• Incident response

What we offer

• Opportunity to work at the forefront of AI research and frontier biology with world‑class colleagues.
• High‑velocity culture with strong ownership and competitive compensation.