Data Scientist (Vice President)

Morgan Stanley Machine Learning (MSML) is Morgan Stanley's center of excellence responsible for working with business and IT teams across the firm to solve mission-critical problems. We are a highly motivated and collaborative team consisting of data scientists, machine learning engineers and members from academia. Our team is uniquely positioned to apply advanced AI to revenue generating business cases.

Responsibilities

- Independently work on end-to-end development of models based on trading data, market data, alternative data and other internal/external data sources

- Lead data science projects and develop models in collaboration with members of MSML team, strats and quants

- Mentor junior members of the team

- Work with stakeholders to refine requirements and communicate progress

- Rapidly prototype and iteratively develop models

- Refine, tune and improve existing models

- Deploy models to production and monitor performance

- Study recent research and develop original ideas to solve hard problems

- Participate in internal and external forums

Qualifications Qualifications

Required

- In-depth expertise in core algorithms such as linear/logistic regression, SVM, random forests, PCA

- Expertise in clustering methods and their applications to high dimensional, large data sets

- Strong command over linear algebra and statistics having the ability to quickly translate ideas to efficient, elegant code

- Development experience in Python (preferred) or Java/Scala with good command over respective data pipelining, matrix algebra and statistics libraries

- Ability to intuitively combine traditional techniques and cutting-edge research to develop novel models

- Some experience with deep learning networks like Feed-forward NN, CNN, RNN and Auto Encoders with Tensorflow or similar libraries

- Experience in model deployment and scaling

- 5-10 years of data science/machine learning experience

- Excellent communication and presentation skills

- MS in Computer Science, Statistics, Financial Engineering or a related quantitative field. PhD preferred.

Desired

- Experience in time series analysis and sequential data using ARIMA, Kalman Filters, HMM, RNN etc.

- Reinforcement learning experience is highly desirable

- Knowledge of parallel computing approaches such as use of GPU parallelization

- Experience in Bayesian Modeling using MCMC, SeqMC and newer techniques like Variational Bayes

- Research publications