Avenews provides fast and reliable financing solutions for Kenyan agribusinesses poised for growth, serving as a trusted partner to fuel expansion.

We're looking for a passionate Machine Learning Engineer or Data Scientist who is excited about building predictive models that have real-world impact in fintech. You'll work on developing and improving credit risk models, particularly probability of default (PD) models, while aggregating and analyzing data from multiple sources. This is a role for someone who loves working with data, is curious about machine learning, and wants to continuously learn and grow.

Responsibilities

• Building and improving probability of default (PD) models for credit assessment
• Aggregating and processing data from multiple sources (MongoDB, SQL databases, APIs, external data providers)
• Designing and implementing end-to-end ML pipelines including data collection, preprocessing, feature engineering, model training, and deployment.
• Developing data pipelines to collect, clean, and transform data for modeling
• Working with financial data and credit scoring systems
• Creating features and performing feature engineering for ML models
• Evaluating model performance and iterating improvements
• Collaborating with the development team to integrate models into production
• Conducting research and experimentation with advanced ML techniques such as ensemble methods, deep learning, and time-series forecasting for credit risk prediction.

Requirements

• 3-5 years of experience in machine learning, data science, or a related quantitative field
• Strong Python experience (this is essential)
• Experience with machine learning frameworks - Required: scikit-learn, XGBoost, LightGBM; Nice to have: TensorFlow, PyTorch, Keras
• Experience with credit risk modeling is a strong plus but not mandatory
• Experience working with multiple data sources and data integration
• Familiarity with APIs - both consuming and building them
• Knowledge of statistical analysis and model evaluation techniques
• Experience with data manipulation libraries (pandas, numpy)
• Understanding of databases - Must-have: MongoDB, PostgreSQL; Nice to have: MySQL, Redis
• Knowledge of data versioning, model versioning, and MLOps pipelines is a plus

• Building and improving probability of default (PD) models for credit assessment
• Aggregating and processing data from multiple sources (MongoDB, SQL databases, APIs, external data providers)
• Designing and implementing end-to-end ML pipelines including data collection, preprocessing, feature engineering, model training, and deployment.
• Developing data pipelines to collect, clean, and transform data for modeling
• Working with financial data and credit scoring systems
• Creating features and performing feature engineering for ML models
• Evaluating model performance and iterating improvements
• Collaborating with the development team to integrate models into production
• Conducting research and experimentation with advanced ML techniques such as ensemble methods, deep learning, and time-series forecasting for credit risk prediction.

• Strong Python experience
• scikit-learn
• XGBoost
• LightGBM
• TensorFlow (Nice to have)
• PyTorch (Nice to have)
• Keras (Nice to have)
• APIs (consuming and building)
• Statistical analysis
• Model evaluation techniques
• pandas
• numpy
• MongoDB
• PostgreSQL
• MySQL (Nice to have)
• Redis (Nice to have)
• Data versioning (plus)
• Model versioning (plus)
• MLOps pipelines (plus)

• BA/BSc/HND
• 3-5 years of experience in machine learning, data science, or a related quantitative field
• Experience with machine learning frameworks
• Experience with credit risk modeling (strong plus but not mandatory)
• Experience working with multiple data sources and data integration
• Understanding of databases
• Knowledge of data versioning, model versioning, and MLOps pipelines is a plus