AI Researcher – Deep Learning for Financial Time Series

Goldman Sachs; New York (NY); Presencial; USD 100 000 - 200 000; Tempo integral

Resumo da oferta

Goldman Sachs AI Research is seeking an experienced Associate to lead design, training, and evaluation of deep learning models for financial time series. You will own end-to-end research problems, from data curation to production-ready architectures, collaborating with quant and engineering teams.

You will work in a fast-moving space, applying rigorous evaluation and architecture choices to deliver robust, generalizable models across desks and asset classes, and contribute to a shared research

Qualificações

• Bachelor’s, Master’s, or Ph.D. in Computer Science, ML, Statistics, Mathematics, Physics, EE, Quantitative Finance, or related field.
• 3+ years of industry experience building, training, and deploying deep learning models, with a focus on sequential or time series data.
• Deep expertise in CNNs, Transformers, autoencoders, GANs, diffusion models, GNNs, Bayesian methods, and reinforcement learning.
• Strong knowledge of classical time series and econometric modelling (ARIMA, GARCH, Kalman filters, state-space).
• Experience with WaveNet, N-BEATS / N-HiTS, DeepAR, PatchTST, TSFMs.
• Expert-level Python and deep proficiency in PyTorch, TensorFlow/Keras and/or JAX/Flax.
• Experience with distributed multi-node GPU training and ONNX workflows.
• Foundations in statistics, probability, stochastic processes, optimization, and signal processing.
• Excellent communication to explain trade-offs to PhD researchers and business stakeholders.
• Comfort operating in ambiguity and rapid change.

Responsabilidades

• Design, implement, and train modern deep learning architectures for forecasting, representation learning, and generative modelling of financial time series (CNNs, Transformers, autoencoders, GANs, diffusion, GNNs, RL).
• Benchmark time-series architectures (WaveNet, N-BEATS/N-HiTS, DeepAR, PatchTST, TSFMs) against econometric methods (ARIMA, GARCH, Kalman).
• Manage large-scale model training across GPU clusters and cloud, including DDP, FSDP, mixed precision, and ONNX deployment.
• Build robust evaluation frameworks for finance data: walk-forward, cross-validation, embargo periods, regime analysis, uncertainty quantification, and backtesting.
• Collaborate with quantitative researchers to translate model outputs into risk-adjusted signals and portfolio insights.
• Contribute reusable models, datasets, benchmarks, and tooling to a firm-wide research platform.
• Mentor junior researchers and ensure model governance, documentation, and auditability.

Conhecimentos

Python

PyTorch

TensorFlow

Keras

JAX

Flax

Distributed training

Time series

Statistics

Communication skills

Formação académica

Bachelor’s/Master’s/Ph.D. in CS/ML/quantitative field

Ferramentas

ONNX

AWS

Kubernetes

Docker

Descrição da oferta de emprego

Goldman Sachs AI Research is seeking an experienced Associate to lead design, training, and evaluation of deep learning models for financial time series. You will own end-to-end research problems, from data curation to production-ready architectures, collaborating with quant and engineering teams.

You will work in a fast-moving space, applying rigorous evaluation and architecture choices to deliver robust, generalizable models across desks and asset classes, and contribute to a shared research