AI Infra Engineer – Real-Time Vision & Multimodal Models

ambient.ai; Redwood City (CA); Hybrid; USD 150,000 - 210,000; Full time

Benefits offered by this job

Stock options

Comprehensive health + welfare package

Flexible time off

Winter Break

Job summary

Ambient.ai in Redwood City is seeking a hands-on infrastructure engineer to design, build, and optimize our AI infrastructure powering real-time security intelligence across video data and sensors.

You will partner with research scientists and product teams to deploy state-of-the-art models, improve inference performance, and build robust evaluation and observability tools for production AI systems. This is a hybrid role with in-office collaboration three days a week.

Qualifications

• Experience designing and building scalable ML infrastructure for training, inference, evaluation, and deployment.
• Hands-on production experience with DL models, including LLMs or multimodal systems.
• Strong knowledge of inference optimization techniques and GPU utilization.

Responsibilities

• Design, build, and maintain AI infrastructure for real-time computer vision, LLM, LVM, and multimodal workloads.
• Scale systems to handle terabytes of video and sensor data with low latency.
• Improve model serving architectures through batching, caching, and quantization.
• Develop evaluation harnesses and benchmarking to ensure production readiness.
• Collaborate with researchers and product teams to bring AI advances into production.

Skills

Python programming

Distributed systems

ML infrastructure

LLMs / LVMs

Inference optimization

GPU utilization

Cloud infrastructure

Containerization

Observability & debugging

Collaboration with researchers

Education

BS/MS in Computer Science or related field

Tools

vLLM

Triton Inference Server

CUDA

NCCL

PyTorch

TensorRT

ONNX

Job description

Ambient.ai in Redwood City is seeking a hands-on infrastructure engineer to design, build, and optimize our AI infrastructure powering real-time security intelligence across video data and sensors.

You will partner with research scientists and product teams to deploy state-of-the-art models, improve inference performance, and build robust evaluation and observability tools for production AI systems. This is a hybrid role with in-office collaboration three days a week.