Founding Multimodal AI Research Engineer

PREDICTION; New York (NY); On-site; USD 180,000 - 240,000; Full time

Job summary

PREDICTION is seeking a Founding AI Research Engineer to help define the AI research direction and build systems for national security. This hands-on role spans video, imagery, text, sensors, and other signals, with opportunities to shape architecture and deployment.

You will work on multimodal reasoning, transformers, model evaluation, and production-ready systems in secure environments. A PhD is helpful but not required; exceptional candidates at various seniority levels are encouraged to

Qualifications

• Strong background in machine learning and deep learning.
• Experience with multimodal learning and video understanding.
• Familiarity with transformers and foundation models.
• Experience with model evaluation, calibration, and robustness.
• Excellent Python and PyTorch proficiency.
• Ability to read research papers and implement ideas.
• Ability to design rigorous experiments.
• Ability to deploy research into production systems.

Responsibilities

• Research and develop new AI capabilities from idea through deployment.
• Design and test multimodal architectures.
• Train, fine-tune, benchmark, and evaluate models.
• Build systems that reason across multiple data types.
• Develop methods for video understanding and temporal reasoning.
• Prototype ideas and turn research into production systems.
• Collaborate with engineers and national security professionals.
• Define PREDICTION's long-term AI research direction.

Skills

Machine learning

Multimodal learning

Video understanding

Transformers

Representation learning

Reinforcement learning

Model evaluation and calibration

Python

Education

PhD (optional)

Tools

PyTorch

Job description

PREDICTION is seeking a Founding AI Research Engineer to help define the AI research direction and build systems for national security. This hands-on role spans video, imagery, text, sensors, and other signals, with opportunities to shape architecture and deployment.

You will work on multimodal reasoning, transformers, model evaluation, and production-ready systems in secure environments. A PhD is helpful but not required; exceptional candidates at various seniority levels are encouraged to