Data Science Manager (Full-Stack / Production ML)

About this role

Overview We are looking for a player-coach who leads a small, high-output data science team while staying deeply hands-on. This role owns the full scope of data science at Appriss Retail — data engineering, governance, and production model delivery — not just model building. The right candidate has built and shipped real data platforms and AI/ML systems using a modern stack, has meaningful experience with LLMs and agentic architectures, and can operate credibly in both the technical weeds and the business conversation.

As a Data Science Manager with 2–4 direct reports from day one, you would set technical direction for the team. This is not a role for someone who manages from a distance. You will write code, review pipelines, define data contracts, and drive architectural decisions — while also growing and directing the team around you. This is a full-stack role, not a model-building role with a handoff. If your last few projects ended when you handed a notebook or a trained model to a data engineering or ML engineering team to put into production, this probably isn’t the right fit.

Essential Du ties Technical leadership & delivery Own end-to-end delivery of high-impact data science projects — from ambiguous business request to production-ready system. Design and maintain data pipelines, data models, and governance standards alongside your team; treat infrastructure as a first-class product concern. Build, evaluate, and iterate on ML models in production; lead experimentation rigor, monitoring, and lifecycle management.

Architect and ship LLM-integrated features and agentic workflows — including prompt engineering, tool use, and output evaluation. Guide cloud infrastructure architecture for data science projects, taking into account performance, maintenance, and cost criteria. Set the standard for code quality: write production-grade Python and SQL, enforce review practices, and maintain documentation. Partner closely with engineering to integrate models and pipelines into core product infrastructure.

People & team Directly manage 2–4 data scientists; provide technical mentorship, career development, and clear performance expectations. Define team operating norms: sprint planning, code review, documentation, and delivery accountability. Recruit and grow the team as the function scales. Strategy & stakeholders Translate ambiguous business problems into well-scoped analytical and modeling work with defined success criteria.