Deep Learning Researcher

Fintal Partners; New York (NY); On-site; USD 180,000 - 250,000; Full time

Job summary

Fintal Partners is seeking a Deep Learning Researcher to build models from the ground up for trading, joining a central research team that partners with traders across the firm.

You will pretrain models from scratch, design novel architectures, and push state-of-the-art methods into production, owning the full data-to-model pipeline in a fast-paced, high-performance environment on-site in New York City or Chicago.

Qualifications

• PhD in CS/Math/Statistics/Engineering or related quantitative field.
• 5+ years building deep learning models with hands-on pretraining or novel architectures.
• Deep expertise in modern deep learning theory and practice.
• Strong Python skills, fluent in PyTorch, TensorFlow, and/or JAX.
• Track record of publications at top venues (ICML, ICLR, NeurIPS, or equivalent).
• Comfortable collaborating in a fast-paced, high-performance team.

Responsibilities

• Architect and train novel deep learning models applied to trading and markets.
• Lead research from first principles, including pretraining models from scratch.
• Push state-of-the-art techniques into production, from idea to deployed model.
• Work closely with researchers, engineers, and traders to turn insights into working systems.
• Own the full data-to-model pipeline, including acquisition, preprocessing, and feature design.

Skills

Deep learning theory

Research leadership

Publications

Education

PhD in CS/Math/Statistics/Engineering

Tools

PyTorch

TensorFlow

JAX

Job description

A global trading firm is looking for a Deep Learning Researcher with a strong record of building models from the ground up to tackle hard problems in trading. This role sits on a central research team working closely with trading groups across the firm. We value researchers who combine deep technical modeling skill with curiosity about markets, and who thrive on solving open-ended problems with rigor and creativity.

What You'll Do:

• Architect and train novel deep learning models applied to trading and markets
• Lead research from first principles, including pretraining models from scratch
• Push state-of-the-art techniques into production, from idea to deployed model
• Work closely with researchers, engineers, and traders to turn insights into working systems
• Own the full data-to-model pipeline, including acquisition, preprocessing, and feature design

What We're Looking For:

• PhD in Computer Science, Math, Statistics, Engineering, or a related quantitative field
• 5+ years building deep learning models, with hands-on experience pretraining models or designing novel architectures (not just fine-tuning or applying existing ones)
• Deep expertise in modern deep learning theory and practice
• Strong Python skills, fluent in PyTorch, TensorFlow, and/or JAX
• Track record of publications at top venues (ICML, ICLR, NeurIPS, or equivalent)
• Comfortable working collaboratively in a fast-paced, high-performance team

This is an opportunity to have significant ownership over a growing ML platform rather than simply maintaining established infrastructure. The team is investing heavily in compute and machine learning, with engineers working directly alongside researchers on systems that have measurable impact in live trading.

Opportunity available on-site in New York City and Chicago.