ML Research Scientist -Deep Learning & Transformer Architectures

What you'll do

• Develop the model architecture, financial-data tokenization scheme, and attention mechanisms.
• Build scalable training, evaluation, and low-latency inference pipelines; conduct ablation studies.
• Collaborate on trading-pipeline integration and document research methods and results.

What they ask for

• PhD in Machine Learning, Computer Science, Statistics, Applied Mathematics, or a related field, focused on deep learning.
• Demonstrated ability to implement Transformer architectures from scratch and deep understanding of attention, positional encodings, tokenization, and training dynamics.
• Expert-level PyTorch; strong Python and C++ skills; mathematical foundations in linear algebra, probability, optimization, and information theory.
• Experience training models at scale (100M+ parameters); ability to define and execute a multi-month research agenda; familiarity with AI-assisted development tools (Cursor, Claude Code).

Nice to have

• Deep learning for financial data or time-series forecasting; tokenization for continuous or non-text data.
• Research published in top ML venues or equivalent industry experience; market microstructure knowledge; model compression, quantization, or inference optimization.

Pay

$150,000 to $200,000 USD base salary

Tools

PyTorch, Python, C++, Cursor, Claude Code

AI focus

Build and train a custom Transformer from scratch for next-token prediction of financial time-series; optimize inference and integrate predictions into live trading.