Data Scientist

We are seeking a Data Scientist to be embedded with our aviation customer, designing and deploying optimisation models that solve real-world constrained resource allocation and scheduling problems in their dynamic operational environment. You will work closely with the customer's operational teams to build demand forecasting models, operationalise them on Databricks using MLflow, and ensure model outputs are explainable and actionable for non-technical stakeholders. This role requires end-to-end ownership from data pipeline development through to delivering transparent, human-accountable decision support systems that keep operational teams fully in control.

Key Responsibilities

• Design, build and deploy constraint-based, scheduling and combinatorial optimisation models that address the customer's resource allocation challenges where demand is dynamic and resources are limited
• Develop time-series forecasting and demand modelling solutions using the customer's historical operational data including passenger volumes, booking patterns and shift requirements, feeding into downstream allocation logic
• Lead the full model lifecycle on Databricks including experiment tracking, model registry, versioning and staged promotion through development, staging and production environments using MLflow
• Create explainable AI outputs and decision support interfaces that enable the customer's operational teams to understand model recommendations and maintain full accountability for their decisions
• Develop and maintain robust data pipelines using Databricks Delta Lake and Medallion architecture patterns, implementing proactive data quality frameworks such as DQX to ensure high data maturity across integrations and real-time systems
• Work directly with the customer's stakeholders to understand operational constraints, validate model performance and iterate on solutions based on real-world feedback
• Establish and document data quality standards and best practices, taking ownership of data preparation rather than handing off responsibility to other teams

Requirements

• Proven direct experience designing and deploying optimisation models such as constraint-based, scheduling or combinatorial approaches in operational environments with dynamic demand and limited resources
• Hands-on experience operationalising models on Databricks with demonstrated competency in experiment tracking, model registry, versioning and environment promotion using MLflow
• Strong track record building time-series forecasting and demand modelling solutions using historical operational data to drive downstream allocation and scheduling decisions
• Experience designing model outputs for non-technical users including explainable recommendations, confidence scores and transparent rationale that operational teams can act upon and understand
• Working knowledge of Databricks Delta Lake, Medallion architecture and data quality frameworks such as DQX, with the ability to build and maintain robust data pipelines rather than treating data preparation as a handoff task

Nice to Have

• Background or familiarity with aviation, airline operations or travel industry use cases
• Experience with real-time and batch data from high-fidelity operational systems and third-party integrations
• Knowledge of other optimisation libraries and frameworks such as Pyomo, PuLP or commercial solvers to complement Databricks-native approaches