Director of Decision Science

At Liberis, our mission is to empower small and medium-sized businesses by removing finance as a friction to growth, delivering contextual, embedded financial solutions to support merchants at every stage of their business lifecycle.

Who are you 

You are a pragmatic leader who combines deep machine learning expertise with commercial credit instincts. You thrive in cross-functional settings, enjoy building high-performing teams, and take responsibility for models in production — from design through validation to monitoring and governance. You are energised by building model solutions that enable partners to deliver great customer experiences while protecting the business. 

This senior leadership role sits at an important inflection point for our product roadmap. You will lead the Decision Sciences function responsible for end-to-end credit model development, validation, and monitoring across a portfolio of B2B embedded-lending products which includes (but not limited to) flexible lines of credit, BCA and its variants. 

What you’ll be doing 

• Lead on model development and delivery: Own end-to-end model lifecycle for approval, scorecards, propensity, fraud, and collections models across seller and buyer products. Design features, run experiments, iterate on multi-bureau data pipelines, and productionize models by working closely with engineering teams. 
• Set modelling strategy and standards: Define modelling standards, validation playbooks, documentation requirements, and SLAs for model deployment and change control. Ensure compliance with regulatory and audit expectations for governance and explainability. 
• Build and scale model monitoring: Implement monitoring for model drift, population stability, performance degradation, and business impact. Create alerting and remediation workflows and own model refresh cadence. 
• Drive technical excellence: Champion advanced ML approaches which include (but not limited to) Gradient boosted decision trees, Survival or Hazard models and Bayesian models. Develop feature engineering and robust statistical techniques for driving SMB lending products. Balance complexity with interpretability and latency constraints. 
• Partner with product, data, and commercial teams: Translate model outputs into decisioning rules, pricing signals, and partner-level policies. Collaborate on experimentation, A/B testing, and propensity-to-convert vs risk trade-offs. 
• Lead, Coach, and Empower the team: Manage a team of 7+ data scientists and analysts including the Head of Decision Analytics. Recruit, mentor, and create clear career trajectories; run technical reviews, code and model clinics, and establish a continuous learning culture. 
• Communicate to stakeholders: Explain modelling choices, promote growth while managing risk trade-offs, and performance to executives, partners, and auditors through crisp written reports and presentations. 

What we think you’ll need 

• Experience: Proven building credit or risk models in financial services with substantial recent experience in B2B unsecured lending for sellers/platforms or embedded-finance ecosystems.
• Technical depth: Proven hands-on experience developing production ML models using XGBoost, GBM, and related techniques. Strong Python and SQL skills for feature engineering, model training, and data validation. Experience with model deployment frameworks and MLOps practices. 
• Full model lifecycle expertise: Demonstrable experience in model design, feature engineering, OOT/OOS testing, validation, calibration, and governance. Familiarity with model explainability, regulatory expectations, and documentation for audit. 
• Product and portfolio thinking: Comfort with approval-rate vs loss-rate trade-offs, renewals economics, pricing impacts, and partner-level P&L. Experience translating model outputs into rules and pricing strategies. 
• Analytics and tooling: Strong data visualisation and analytics skills (Power BI, Tableau, or equivalent). Able to prototype quickly and collaborate with engineering to productionize models. 
• Judgment and communication: Track record making high-impact decisions on large exposures with clear rationales and reproducible audit trails. Excellent written and verbal communication with senior stakeholders. 
• Leadership: Prior people management or function-lead experience with evidence of building standards, processes, and a culture of continuous improvement. 

Nice to have

• Prior experience with e-commerce or ISV partnership models in the US and/or UK 
• Hands on experience with decision science libraries such as scikit-learn, XGBoost/LightGBM/CatBoost, statsmodels; familiarity with SHAP/LIME for explainability. 
• Exposure to commercial bureaus and third-party data vendors (Experian, Equifax, D&B) and alternative data sources common in seller ecosystems. 
• Working knowledge of Monitoring & quality tools such as Evidently AI, WhyLabs, or equivalent for drift/PSI; Great Expectations for data tests. 
• General familiarity with ML platforms & MLOps: AWS SageMaker (Studio, Pipelines), Databricks ML, MLflow, Feature Store, Docker/Kubernetes for serving, CI/CD (GitHub/GitLab), orchestration (Airflow/Prefect), and IaC where relevant. 
• Working knowledge of implementing Agentic AI solutions at scale 

What happens next?

Think this sounds like the right next move for you? Or if you’re not completely confident that you fit our exact criteria, apply anyway and we can arrange a call to see if the role is fit for you. Humility is a wonderful thing, and we are interested in hearing about what you can add to Liberis!

Our hybrid approach

Working together in person helps us move faster, collaborate better, and build a great Liberis culture. Our hybrid working policy requires team members to be in the office at least 3 days a week, but ideally 4 days. At Liberis, we embrace flexibility as a core part of our culture, while also valuing the importance of the time our teams spend together in the office.

 #LI-CG1