Predictive Analyst

Predictive Analyst

About The Role

Key Responsibilities

• **Data Strategy & Preparation:** Review and identify key datasets from partners to develop robust ML models. Collect, preprocess, and analyze large, complex datasets to ensure data integrity.
• **Feature Engineering:** Define features and perform advanced clustering, feature reduction, and selection techniques.
• **Model Development:** Build predictive models to improve partner and product performance, developing specific "Ezra Prediction" modules in Python for automated ML systems.
• **Experimental Design:** Design and conduct experiments (A/B testing) to validate models and hypotheses, enhancing overall market understanding.
• **Deployment & Collaboration:** Partner with cross-functional teams to implement and deploy machine learning solutions into production environments.
• **Insight Communication:** Translate complex findings into actionable insights for both technical and non-technical stakeholders through reports and visualizations.
• **Optimization:** Continuously monitor and iterate on model performance, prediction systems, and underlying algorithms.

Essential Qualifications And Experience

• **Education:** Bachelor's or Master's degree in Computer Science, Statistics, Mathematics, or a related quantitative field.
• **Professional Experience:** 5+ years of experience in a Data / Decision Science or Machine Learning Engineering role with a proven track record.
• **Technical Proficiency:**
• + Expert-level Python programming skills.
• + Advanced SQL skills (Postgres, Redshift, or similar).
• + Deep understanding of data structures (lists, dictionaries, arrays, data frames).
• + Data visualization / analysis / statistical inference.
• **Core Competencies:** Strong analytical and problem-solving skills, with the ability to work effectively both independently and within a team.
• **Communication:** Exceptional story-telling and presentation skills, specifically the ability to bridge the gap between technical complexity and business logic.

Beneficial Qualifications And Experience

• **MLOps & Production:** Experience operationalizing ML models into production with measurable impact on product and financial optimization.
• **DevOps Tools:** Proficiency with Source Control Management (Git/Bitbucket) and CI/CD pipelines (e.g., Jenkins).
• **Architecture:** Understanding of Feature Store architecture and implementation.
• **Environment:** Experience working in Linux OS variants (Debian) and comfort utilizing the command line.
• **Agility:** Ability to balance "start-up" stage project work with the development of long-term scalable ML modules.
• **Language Skills:** Proficiency in **French** is beneficial to support operations across various jurisdictions.

The Ideal Candidate

• **Innovation-Driven:** Passionate about deploying customer-centric products that exceed quality and delivery expectations.
• **Growth Mindset:** Proud of their work but driven by continuous improvement; able to self-critique and accept constructive feedback.
• **Detail-Oriented:** High attention to detail with a commitment to process, procedure, and documentation as pillars of excellence.
• **Adaptable:** Comfortable stepping outside of their comfort zone to learn new tasks and technologies.

Company Overview

• Ezra provides B2B digital lending solutions for emerging markets in partnership with mobile and digital wallet operators and financial service providers.
• Ezra supports 24 operations in 23 countries, across Africa, the Middle East and Asia.
• Our key office locations are in Nairobi, Kenya and Dubai, UAE.
• Our flagship products are Airtime Credit Services (ACS), Nano and BNPL.
• ACS is an airtime or data advance offered to prepaid mobile subscribers at the point of low credit.
• Nano is a micro cash advance offered to mobile wallet users on demand.
• BNPL facilitates payment installments for products and services
• As a FinTech company, our business is entirely technology and data driven, from determining subscriber eligibility, generating relevant offers, managing risk, loan issuance, recovery, optimizing performance and reporting, reconciliation and billing.
• Each day we process approximately 21M loan requests and 1.4 TB of data across our markets.
• This process needs to be robust, reliable and secure.
• But it doesn't end there.
• We're exploring new ways of using our platform and transactional data to improve our products and develop new product opportunities.