Real-Time Vision & Multimodal AI Engineer

Everseen; Herndon (VA); On-site; USD 110,000 - 150,000; Full time

Job summary

Everseen is seeking a hands-on ML/vision engineer to design, train, and productionize vision and multimodal models powering our next-generation platform. You will ship models that run on live retail video at scale, with latency-aware deployment and GPU acceleration.

You will collaborate with engineering, research, and MLOps to apply TensorRT, ONNX, PyTorch, and related tooling to optimize real-time performance across thousands of stores and checkouts.

Qualifications

• 3–5 years building and shipping ML/vision systems with real-time production exposure.
• Experience with vision-language models (VLMs) and world models; know where they help and where they don’t.
• Strong Python programming and hands-on experience with PyTorch.

Responsibilities

• Design, train, evaluate and own vision and multimodal models from prototype to production.
• Apply VLMs and world models to retail problems at scale and optimize for latency and cost.
• Build and optimize real-time vision AI pipelines processing live video across thousands of stores.
• Collaborate with research and platform engineers to deploy new architectures under tight constraints.
• Run experiments, measure results, and ship based on data.

Skills

Python

PyTorch

VLMs

World models

Real-time systems

CUDA

CI/CD

Git

Education

Bachelor's or Master's in CS/ML/Engineering

Tools

Linux

Docker

Kubernetes

TensorRT

ONNX

CUDA

Job description

Everseen is seeking a hands-on ML/vision engineer to design, train, and productionize vision and multimodal models powering our next-generation platform. You will ship models that run on live retail video at scale, with latency-aware deployment and GPU acceleration.

You will collaborate with engineering, research, and MLOps to apply TensorRT, ONNX, PyTorch, and related tooling to optimize real-time performance across thousands of stores and checkouts.