Founding Applied Scientist - Production Computer Vision

Visa Hunt; United States; On-site; USD 180,000 - 280,000; Full time

Job summary

Visa Hunt is assisting an innovative AI-driven retail security firm experiencing rapid growth. They seek a Founding Applied Scientist to build core AI capabilities from the ground up, owning the full model development lifecycle from data collection through training, optimization, and shipping production-quality code.

You will work with the CTO and a distributed engineering team to deploy computer vision solutions for real-world retail safety and operations, including YOLO-style architectures in

Qualifications

• 3+ years of professional experience in Applied Science, ML Engineering, or Computer Vision roles.
• Shipped customer-facing products with real-world CV applications (e.g., YOLO or similar).
• Advanced proficiency in Python with PyTorch or TensorFlow.
• Master’s degree or PhD in Computer Science, ML, or CV.
• Collaborative startup mindset, across time zones (Turkey-based engineering mentioned).

Responsibilities

• Lead model lifecycle including production training, optimization, and inference deployment.
• Write production-ready Python code to deploy CV models with cross-functional teams.
• Design data collection and quality assessment frameworks to improve model accuracy.
• Present technical findings and AI roadmaps to engineering teams and leadership.

Skills

Python

PyTorch

TensorFlow

Computer Vision

Production-grade code

Education

Master’s degree in Computer Science

PhD in Computer Science/ML/CV

Tools

YOLO family

TorchScript

Job description

Visa Hunt is assisting an innovative AI-driven retail security firm experiencing rapid growth. They seek a Founding Applied Scientist to build core AI capabilities from the ground up, owning the full model development lifecycle from data collection through training, optimization, and shipping production-quality code.

You will work with the CTO and a distributed engineering team to deploy computer vision solutions for real-world retail safety and operations, including YOLO-style architectures in