# 7 - Senior AI Data Scientist / Modeler - Databricks - Mexico

Muttdata

1d•Remote

About The Position

Muttdata is hiring a Senior AI Data Scientist / Modeler with Databricks experience for a project with a leading multinational beverage company based in Mexico City. The role will build the Sales Leader Insights product within the Smart Force platform, an agentic AI architecture that selects relevant insights for each sales leader and recommends a practical coaching route, balancing geographic proximity with route priority and criticality. The position combines agentic AI, recommendation systems, and operations research. It works closely with data, platform, and business teams throughout the AI Factory cycle, from governed data and features through model development, orchestration, and productionization. Strong technical depth, business sense, and ownership are important for success. The role is remote-first; Mexico City (Santa Fe / Cuajimalpa) is preferred, but the position is not strictly onsite and other locations are welcome.

Requirements

Proven experience as a Data Scientist or Modeler in supervised machine learning and recommendation or prioritization systems.
Hands-on experience developing agentic AI, including designing decision-making AI agents.
Experience in optimization or operations research, including routing, vehicle routing problems (VRP), or constrained optimization.
Advanced Python skills, including scikit-learn and optimization libraries such as OR-Tools, Pyomo, or PuLP, or similar.
Experience with MLflow and model development in Databricks.
Experience working with geospatial or routing data, including distances and geolocation.

Nice To Haves

Previous experience with sales forces, commercial coaching, or field/route optimization solutions.
Experience developing AI agents or agentic infrastructure, such as Mosaic AI Agent Framework, agent orchestration, or MCP.
Familiarity with Unity Catalog, Databricks Asset Bundles, and Azure DevOps.

Responsibilities

Design and build an agentic AI architecture that generates and prioritizes actionable insights for sales leaders based on management focus, coaching needs, and performance.
Model coaching-route recommendations as a routing optimization problem, balancing geographic proximity with route priority and criticality while avoiding inefficient trips.
Integrate business signals, including potential, execution, compliance, and pre-seller KPIs, so coaching recommendations are operationally and geographically practical.
Enable agentic decision-making in which an AI agent evaluates and adjusts recommendations according to time, distance, and criticality constraints.
Package the model under the MFL framework, including PyFunc, a model card, and tests, and integrate it with orchestration and deployment to operations.