Data Scientist at Investec

Role Overview

• The primary expectation of this role is to solve business problems through the extraction, analysis and interpretation of data using algorithmic, statistical and machine learning tools, in order to develop models that understand patterns and predict useful business outcomes (eg. Product and next best action recommendation, behavioural and lifestyle clustering, client lifetime value, etc). Combining this with data storytelling and effective communication skills, the ask is to quantify business value and support decision making by delivering a data story to the business in a meaningful and understandable way.
• Our operating model is built around agency and autonomy, asking our team members to build end-to-end partnerships and drive solutions to completion as per the business need. From engaging stakeholders on the business requirement, to collaborating on project prioritisation, to developing the technical insights and solutions, to presenting the findings through data storytelling; you will own and drive projects while working with relevant internal and external team members.  Coupled with generating innovative ideas for new models, a risk-conscious approach to toolset and dataset selection as well as overall solution design is critical, along with a firm view to ensure fairness and minimise bias in model outputs.
• Adherence to important regulatory standards such as POPIA is a must.

Key Responsibilities

• Gather data from structured and unstructured sources, whether internal or external, and clean and preprocess data to ensure quality and consistency
• Utilise machine learning algorithms to design, develop, test, and review predictive and prescriptive models that align with and support our business objectives, using both on-premise and cloud-based infrastructure
• Analyse and interpret data, trends and patterns and deliver insights, data stories and recommendations to enhance business strategy, both independently and collaboratively
• In addition to model accuracy and selecting fit-for-purpose tools, ensure that bias mitigation and ethical considerations are cornerstones of model development
• Develop graphs, dashboards, and presentations of project results and present to key stakeholders
• Collaborate with ML Ops engineers to prepare models for productionisation in the cloud
• Use feedback loops and general analysis to improve model performance over time
• Offer specialised data science expertise and introduce new ML techniques where appropriate
• Proactively manage projects for timely and accurate completion within scope of responsibility
• Engage with key stakeholders to improve, deliver, pivot and review strategic initiatives
• Develop new ML use cases and quantify their commercial viability
• Engage in prioritisation discussions for projects to maximise commercial value and ensure alignment with strategic goals
• Enhance model development processes to drive automation and reduce manual tasks
• Evaluate current tools and technologies to develop use cases for upgrades and enhancements
• Contribute to the broader data science community within Investec to share knowledge, collaborate in problem solving, drive tool usage, enhance processes, and forge new relationships
• Mentor junior data scientists
• Assist in enhancing data science literacy within the organisation
• Collaborate to enhance our model governance frameworks and ensure these are applied to the building of advanced analytics solutions
• Maintain the integrity of data processes to ensure continuous improvement of data quality that supports compliance with legal, regulatory and industry best practice
• Comply with security and audit controls to protect data solutions and their environment
• Keep abreast of latest developments in data science, data, technology, banking and global events

Minimum Qualifications and Knowledge

• A postgraduate degree in data science or a field related to data science, such as Computer Science, Statistics, Mathematics, Engineering, etc
• 3-5 years of experience in development, testing, validation and monitoring of machine learning models
• Evidence of experience with data-driven problem solving and statistical analysis (descriptive and inferential)
• Experience with deploying ML models in production, including an understanding of ML Ops principles and best practices
• Proficient in Python, SQL, and Jupyter notebooks (PySpark is beneficial)
• Competent in visualisation tools (eg. PowerBI) and Microsoft Office (Excel, Powerpoint)
• Preferable: Experience in DevOps practices, version control, etc
• Preferable: Experience in cloud platforms and tools such as Microsoft Azure, including Azure ML, Azure Dev Ops (ADO), Azure ML Feature Store and Databricks

Competencies

• Ability to develop, test, and optimise machine learning models
• Understanding of how models deliver business value required to advance strategy
• Strong analytical and critical thinking skills
• Ability to collaborate effectively with cross-functional teams
• Excellent verbal and written communication skills
• Inquisitive mindset
• Ability to connect solutions to their commercial impact
• A focus on ethical considerations in data science (bias, fairness, etc)
• Comfort in iterative delivery
• Results orientated, producing a high standard of work
• Ability to work under time pressures on multiple projects
• Attention to detail
• Self-starter – must be proactive and productive with minimal direction
• Ability to work in a fast-paced, technical, cross-functional environment