Fondery |Data Scientist at Rand Merchant Bank

Job description

What we do

In the Foundery we take problems, innovate them into large solutions and start building in small steps. We like to think big, start small, scale quickly.

Purpose

To build and support an experimental data science capability in RMB FOUNDeRY. This team focuses on exploring the application of data science techniques across all aspects of corporate and investment banking. At the FOUNDeRY we take problems, innovate them into large solutions and start building in small steps. We like to think big, start small, scale quickly.

Academic requirements

• Academic Qualifications: Preferably an Honours or Masters degree in a scientific discipline such as:
• Mathematics or Applied Mathematics;
• Computer Science;
• Statistics;
• Engineering; and
• Physics.

This requirement may be relaxed if there is evidence that the candidate has worked in a data scientist role for minimum of 2 years.

Other requirements

• Thrives in an environment of ambiguity and competing priorities
• Demonstrates strong partnering and collaboration skills
• Proficiency with visual analytics and storytelling through visualizations
• Must be a creative, critical thinker who enjoys problem solving
• High aptitude and passion to learn new science, software applications, and tools
• Strong written, communication, and presentation skills

Work Experience

Minimum: 3+ years in a quantitative/analytical role. Preferably, the candidate should have worked in an environment where results are communicated regularly.

Skills:

• Machine Learning Knowledge: An advanced understanding of the machine learning. The candidate must have the ability to clearly explain concepts such as:
• Supervised Learning
• Regularisation and Cross-Validation;
• Types of classification algorithms;
• Methods for parameter configuration/estimation;
• Metrics for measuring model performance;
• Class Imbalance; and
• Dimensionality reduction and Feature transformation.
• Unsupervised Learning
• Clustering;
• Anomaly Detection;
• Metrics of similarity;
• Techniques for handling incomplete or missing information;
• Optional: Reinforcement Learning & Natural Language Processing.
• Programming Skills: The candidate must be competent programmer who is quick to learn new languages.
• The candidate must show that they have previously built statistical learning solutions/products using Python, R, Matlab or other appropriate language.
• Python experience is preferred.
• Experience with R and the use of libraries and packages (such as caret) would be advantageous.
• Big data and cloud platforms: There are a wide range of other tools/platforms which may have utilised specifically for very large datasets to deploy solutions which have machine learning components, such as: Spark, Hadoop, MongoDB, Microsoft Azure etc. Knowledge of such systems would be advantageous.
• Version Control System: Familiarity with source/version control systems such as Git, GitHub and GitLab is required.
• Databases: Must have experience with relational databases such as MS SQL Server and PostgreSQL. The candidate must have the ability to setup processes which curate, collate and store information (possibly unstructured) in the database from a variety of sources. They must be able to understand indexes, views and stored procedures.
• Experience with developing and deploying recommendation algorithms; preferably with real-time inputs and location-based awareness
• Hands-on experience with real-time or micro-batch predictive analytics using machine learning

Responsibilities

The candidate must be able to:

• Compile and aggregate data sets from multiple, disparate sources;
• Handle and pre-process large unstructured and structured datasets;
• Rapidly prototype hypothesis and models;
• Schedule and update production-ready models;
• Maintain and improve data science components in the system, i.e., improve the execution time of existing models, make amendments to existing code base to increase the model performance;
• Produce and maintain documentation of machine learning models and processes;
• Apply data mining, quantitative analysis, and statistical models to identify non-obvious patterns in data;
• Prescribe analytic techniques that produce impactful results, as well as describe those techniques to a diverse set of stakeholders.

Personal Attributes

• Passion for data and what can be done with it
• Attention to detail with a focus on accuracy and quality
• Ability to work in a dynamic environment
• Ability to deliver quality work under a high amount of pressure;
• Keeping up to date with latest trends in data science research and technologies;