This list contains only the countries for which job offers have been published in the selected language (e.g., in the French version, only job offers written in French are displayed, and in the English version, only those in English).
The CNPF Data & AI organisation is looking for a Director of Data Engineering (L5) to lead the strategy, architecture, and delivery of the data platform powering our analytics products and agentic AI applications across Small & Medium Enterprise (SME), Corporate Solution, Transfer Solution and Commercial Verticals. This is a senior, hands-on technical leadership role within Data & AI Product Enablement. The Director will own the data backbone that our LLM agents, MCP servers, and analytics products run on — making sure data is reliable, governed, retrievable in real time, and ready for AI consumption at production scale. The role works in close partnership with Applied AI, Product, and Architecture leadership.
Job Responsibility:
Own the data engineering strategy and technical direction for CNPF, with a strong focus on enabling agentic AI and GenAI products in production
Architect and deliver the data foundations for multi-agent systems — including MCP servers exposing data and tools to agents, retrieval pipelines, vector stores, feature stores, and knowledge graphs
Lead the design of context-engineering infrastructure that lets agents reason over Mastercard data safely, with the right grounding, freshness, and access controls
Drive lakehouse, streaming, and event-driven platform design (Databricks, Spark, Kafka, Delta/Iceberg) to support both batch analytics and low-latency AI use cases
Ensure data systems meet Mastercard standards for governance, lineage, data quality, observability, and risk — including the additional requirements that come with AI consumption (PII handling, prompt/response logging, audit trails)
Set technical standards for how data products are exposed to agents and applications, including MCP design patterns, schema contracts, and tool interfaces
Partner with Applied AI on evaluation and runtime data needs — training sets, eval datasets, retrieval quality, and feedback loops
Stay hands-on enough to make sharp architectural calls, review designs, and unblock the team on hard problems
Guide a team of senior data engineers, providing technical direction and growing their capability over time
Requirements:
Significant experience leading the design and delivery of large-scale data platforms in production
Deep expertise in distributed data processing and the modern data stack — Spark, Databricks, Kafka, dbt, Delta/Iceberg, and similar
Strong hands-on background in data architecture, modelling, streaming, and lakehouse design on AWS
Proven track record of taking data systems from concept to secure, scalable production
Solid grasp of data governance, lineage, quality, and observability frameworks
Excellent technical communication — able to align engineers, AI scientists, product managers, and executives
Comfortable operating as a player-coach: setting direction, reviewing designs, and going deep when needed
Nice to have:
You have personally built data infrastructure that powers agentic AI in production — not just analytics dashboards
Hands-on experience designing and operating MCP (Model Context Protocol) servers, including authentication, tool exposure, schema design, and observability
Direct experience building the data layer for multi-agent systems — retrieval, memory, state management, long-running workflow data, and human-in-the-loop checkpoints
Strong familiarity with vector databases, hybrid retrieval (semantic + structured), and knowledge graph integration with LLMs
Practical understanding of LLMOps data needs — eval datasets, golden traces, prompt/response telemetry, and feedback capture
Experience designing real-time and event-driven systems that support low-latency agent decisioning
Sharp instincts for the trade-offs between batch and streaming, structured and unstructured, accuracy and cost — and how those decisions cascade into agent behaviour
Experience partnering with security and governance teams to ship AI-facing data products responsibly at enterprise scale