Applied AI Research Engineer

Beam, an ultrafast AI inference platform backed by Y Combinator and Tiger Global, is hiring an Applied AI Research Engineer. The role involves owning inference research to lower cost per token and latency, including low-level optimization techniques like speculative decoding, quantization, and KV-cache management. You will work directly with customers and have high autonomy to drive platform improvements.

• Own inference research hands-on.
• Find ways to lower cost per token and latency on customer workloads.
• Low-level inference optimization (speculative decoding, quantization, KV-cache, memory management).
• Work directly with customers to optimize their production workloads.
• Apply learnings to platform as product improvements.
• High-level autonomy to identify high-impact opportunities and guide platform direction.

• Systems or research background in LLM inference.
• Deep understanding of LLM serving, from the kernel to the scheduler.
• History of shipping products or research that people use in production-like scenarios, whether academic or industry.
• Excited to collaborate closely with customers.
• Enthusiasm for developer tools, cloud native technologies, and open source software.