AI Quant Trader & Research Scientist (Foundation Models)

Next Frontier Capital; Singapore; Sur place; SGD 180 000 - 240 000; Plein temps

Résumé du poste

J.P. Morgan is seeking an AI/ML quantitative researcher to pre-train large foundation models from scratch. You will work on Transformer-based and time-series models across FX, rates, commodities, and more, targeting robust, transferable outcomes for live trading environments.

Responsibilities include developing representations, training recipes, and evaluation protocols that connect metrics to economic performance, while collaborating with ML infrastructure engineers on scalable pipelines.

Qualifications

• Advanced degree in ML, CS, statistics, math, OR engineering or related field.
• At least 2 years of relevant experience in ML or quantitative research.
• Proven experience pre-training a large model from scratch beyond API use.
• Experience building large-scale data pipelines and distributed training.
• Deep knowledge of large-model training, optimization, and benchmarking.

Responsabilités

• Pre-train Transformer-based and time-series foundation models from scratch on market data.
• Develop data representations, tokenization, self-supervised objectives, and training recipes.
• Fine-tune and post-train models for alpha generation, pricing, and risk tasks.
• Study scaling laws, transfer across instruments, and robustness with latency trade-offs.
• Design evaluation protocols linking pre-training to economic outcomes.
• Build reusable training, evaluation, and model-serving components with ML infra engineers.

Connaissances

Transformer models

Time-series modeling

Large-scale training

PyTorch

JAX

Distributed training

Research publications

Formation

Advanced degree (Master’s/PhD)

Equivalent experience

Outils

PyTorch

JAX

Description du poste

J.P. Morgan is seeking an AI/ML quantitative researcher to pre-train large foundation models from scratch. You will work on Transformer-based and time-series models across FX, rates, commodities, and more, targeting robust, transferable outcomes for live trading environments.

Responsibilities include developing representations, training recipes, and evaluation protocols that connect metrics to economic performance, while collaborating with ML infrastructure engineers on scalable pipelines.