Engineering Manager, Model Inference

Abridge Abridge · Vertical AI · San Francisco, CA · Builder

Engineering Manager to lead and grow the Model Inference team, focusing on architecting and scaling low-latency, high-throughput inference infrastructure for Abridge's generative AI products in healthcare. Responsibilities include technical direction, model optimization, deployment, and team leadership.

What you'd actually do

  1. Lead and grow a high-performing team of AI inference engineers focused on building and scaling infrastructure for Abridge’s products and APIs
  2. Own the technical direction of our inference systems—making key decisions around batching, throughput, latency, and GPU utilization
  3. Architect and scale inference infrastructure for reliability, efficiency, and observability; lead incident response
  4. Benchmark and eliminate bottlenecks throughout the inference stack
  5. Partner with ML Research teams on model optimization, quantization, and deployment

Skills

Required

  • 5+ years of engineering experience
  • 1+ years in a technical leadership or management role
  • Deep, hands-on experience with ML systems and inference frameworks (e.g., PyTorch, TensorRT, vLLM, TensorFlow)
  • Strong understanding of LLM architecture (eg. Multi-Head Attention, Multi/Grouped-Query Attention, and common transformer components)
  • Experience with inference optimizations (eg. batching, quantization, kernel fusion, FlashAttention)
  • Familiarity with GPU characteristics, roofline models, and performance analysis
  • Experience deploying reliable, distributed, real-time systems at scale
  • Experience with parallelism strategies: tensor parallelism, pipeline parallelism, expert parallelism
  • Skilled at hiring and mentorship
  • Strong technical communication and cross-functional collaboration skills
  • Comfortable giving constructive feedback on technical designs and code reviews
  • Has thrived in a fast-growing startup and knows how to operate with urgency and focus

Nice to have

  • Background in training infrastructure and RL workloads
  • Skilled in building secure, compliant systems on major cloud platforms (GCP preferred, AWS experience welcome)
  • Experience with Kubernetes and container orchestration at scale
  • Published work or contributions to inference optimization research

What the JD emphasized

  • low-latency, high-throughput infrastructure
  • LLM serving techniques
  • AI inference engineers
  • batching, throughput, latency, and GPU utilization
  • inference infrastructure
  • model optimization, quantization, and deployment
  • inference optimizations
  • parallelism strategies

Other signals

  • Owns the end-to-end technical direction of how our models are served
  • architecting low-latency, high-throughput infrastructure
  • pushing the frontier of LLM serving techniques
  • lead a high-performing team of AI inference engineers
  • partner closely with ML Research
  • ensure the systems underpinning every clinician interaction are operating at peak efficiency and reliability