ML Research Scientist -Deep Learning & Transformer Architectures

ML Research Scientist -Deep Learning & Transformer Architectures

Please direct all resume submissions to [email protected] and reference REQ-29605 in the subject.

Overview 
As part of a long-term research agenda within a newly formed systematic equities pod, we are building a proprietary Transformer-based model trained on tokenized intraday market data for next-token prediction of price movements.

We are seeking an exceptional ML research scientist with deep expertise in Transformer architectures and large-scale model training. You will design, implement, and train a custom decoder-only Transformer from scratch -not fine-tune an existing LLM, but build a purpose• built architecture for financial time-series.

This is a long-term research project with significant computational resources. The successful candidate will have a PhD in machine learning or a related field and demonstrated ability to implement Transformer architectures from first principles.

Principal Responsibilities

•    Design and implement a custom decoder-only Transformer architecture optimized for tokenized financial time-series data
•    Develop a novel tokenization scheme for intraday market data: price movements, volume, order flow, and cross-sectional features
•    Implement efficient training pipelines using PyTorch with mixed-precision training, gradient checkpointing, and multi-GPU parallelism
•    Design attention mechanisms adapted to financial data: temporal attention patterns, cross-asset attention, and multi-scale representations
•    Build evaluation frameworks for next-token prediction accuracy, signal quality, and trading performance
•    Implement inference optimization for low-latency production deployment: model quantization, KV-cache, speculative decoding
•    Conduct rigorous ablation studies to validate architecture choices and training methodology
•    Collaborate with the team to integrate model predictions into the live trading pipeline
•    Document research methodology, experimental results, and architectural decisions

Required Skills / Qualifications

•   PhD in Machine Learning, Computer Science, Statistics, Applied Mathematics, or a related field with a focus on deep learning

•    Demonstrated ability to implement Transformer architectures from scratch (not just fine­tuning pre-trained models)
•    Deep understanding of attention mechanisms, positional encodings, tokenization strategies, and training dynamics
•    Expert-level PyTorch skills including custom modules, training loops, mixed-precision, and multi-GPU training
•    Strong mathematical foundations: linear algebra, probability theory, optimization, information theory
•    Experience training models at scale (100M+ parameters)
•    Strong programming skills in Python and C++ for performance-critical components
•    Self-directed researcher capable of defining and executing a multi-month research agenda
•    Familiarity with Al-assisted development tools (Cursor, Claude Code)

Preferred Skills / Experience

•    Experience applying deep learning to financial data or time-series forecasting
•    Familiarity with tokenizatlon approaches for continuous or non-text data
•    Published research in top ML venues (NeurlPS, ICML, ICLR) or equivalent industry experience
•    Knowledge of market microstructure and intraday trading dynamics
•    Experience with model compression, quantization, and inference optimization