AI Engineer - Real-Time Multimodal Systems (On-site NY)

Twin Prime; New York (NY); Sur place; USD 110 000 - 190 000; Plein temps

Résumé du poste

Twin Prime in New York is seeking an AI Engineer to bridge research and production, building reliable systems that process real‑time sensor data. You will implement new modeling approaches, create training and evaluation pipelines, and push models into production in collaboration with researchers and software engineers.

The role emphasizes hands‑on ML development across model life cycle, optimizing performance, and deploying multimodal, vision‑driven models on constrained hardware.

Qualifications

• Hands-on experience training, fine-tuning, evaluating, and deploying deep learning models in research or production.
• Strong programming skills in Python and PyTorch or comparable frameworks.
• Solid understanding of deep learning fundamentals, architectures, optimization, and evaluation.
• Experience in computer vision, multimodal learning, or generative modeling.
• Ability to translate research ideas into working, maintainable code.
• Ability to diagnose problems across data, model, and system performance.

Responsabilités

• Implement and adapt models for video understanding, spatial reasoning, and multimodal learning.
• Build data processing, training, and evaluation pipelines.
• Design experiments, analyze model behavior, and translate findings into improvements.
• Optimize inference performance, memory usage, and deployment on constrained hardware.
• Integrate models into reliable systems that process complex sensor data in real time.
• Build tools to make research workflows reproducible and accelerate path from prototype to deployment.

Connaissances

Python

PyTorch

DL Fundamentals

Computer Vision

Multimodal Learning

Software Engineering

Experimentation

Systems Debugging

Outils

CUDA

C++

Description du poste

Twin Prime in New York is seeking an AI Engineer to bridge research and production, building reliable systems that process real‑time sensor data. You will implement new modeling approaches, create training and evaluation pipelines, and push models into production in collaboration with researchers and software engineers.

The role emphasizes hands‑on ML development across model life cycle, optimizing performance, and deploying multimodal, vision‑driven models on constrained hardware.