Senior Machine Learning Engineer - Fraud

Your key area of focus:

The main focus of this role will be within the FinCrime teams, focusing on innovative ways to detect bad actors while enhancing the overall user experience. As the role develops there will be other areas of the business that will also need support ie. risk and credit rating

You will champion the data science space, looking for where there are the greatest opportunities to deploy models/algorithms and find the ‘low hanging fruit’.

What you will own:

• Own our existing fraud detection models

• Evaluate and integrate new data sources for our algorithms, aligning with Data Engineering and Analytical Engineers' best practices for dbt

• Automate the training and deployment of updated models, ensuring the output is tested, automated, scalable and documented and checks are in place to identify drift.

• Help build experiments to evaluate new models, third-party data sources and tooling.

• Translate commercial requirements into technical solutions, converting real-world problems into solvable data science projects, resulting in insights that further the strategy and enable visibility into key results

• Improving existing models through greater scrutiny of the methodology and improving the input data 

• Develop strategies and tools to help less technical individuals understand and use the models and results.

Who you are:

• You are a problem solver who can identify opportunities for data-driven solutions and prioritize against commercial impact

• You are motivated to deeply understand user behaviour and deliver actionable recommendations to teams alongside a strong technical data solution.

• You can confidently discuss complex business and technical topics with a range of stakeholders and present findings

What you bring to the table:

• 4+ years of professional experience training and deploying models that deliver measurable value (regression, clustering, decision trees, cost-sensitive Machine Learning etc with an emphasis on gradient boosting-based methods).

• You have strong SQL skills, confidently able to pull and manipulate data to get into the desired format for modelling (CTEs, joins, case statements, subqueries)

• Possess strong Python skills, able to automate processes and deploy applications. you are able to deploy your stuff and be able to set up at least basic monitoring.  

• Familiar with building and deploying web applications using Python web frameworks.

• Experience in one or more of the following areas:

  • Machine Learning (Scikit Learn,  XGBoost, H2O etc...)

  • SQL Analytics (BigQuery, Redshift, Databricks, Athena, etc)

  • Visualisation Tools (matlibplot, seaborn, streamlit Looker, Tableau, Periscope, etc)

Bonus points if you

• You have experience with graph databases

• Have experience working with cloud-based services, especially AWS(e.g. Sagemaker, ECS, EMR)

• Have experience with experimentation design and evaluation

• Demonstrate tenacity and a willingness to go the distance to get something done. You don't mind doing things manually but automate at every opportunity.

• Are inquisitive, intellectually curious and can make sense of complex systems or information.

• Can work in a structured approach towards goals and pay attention to detail.

• Can easily communicate with non-technical folks and translate their feedback into code.

• Are comfortable defaulting to over-communication and overreaching when it comes to coordination

• Adjust quickly to changing priorities and conditions and cope effectively with complexity and change.

Key details

• Team Composition: The FinCrime Team is a combination of Data Scientists, Machine Learning Engineers and Backend Engineers

• Location: Our company is remote first. You can be based anywhere in Africa, Europe, or the Americas

• Length of position: Permanent.