Director, Data Architecture

Select how often (in days) to receive an alert:

At Phillips 66, we care about the whole person. That means providing access to quality programs, resources and care that support you and your family. We believe well-being — physical, emotional, social and financial — plays an important role in helping our people thrive. When our employees are supported at every stage of life, they’re empowered to do their best work and make a meaningful impact, both personally and professionally.

As the Director, Data Architecture
you will be the founding architect for the commercial knowledge layer. It is a principal-level role that owns the semantic and data architecture unifying market, operational, and trading data into governed, reusable data products that traders and data scientists can trust. The value here is the trading-specific model of the data, not the infrastructure beneath it: you will partner with Enterprise Analytics' Data-as-a-Service platform and build on it rather than rebuild it.

Phillips 66 Commercial is standing up its data science and analytics capability, and this is a rare chance to design its data architecture from the ground up and own it end to end. Traders, market analysts, and data scientists are ready to move faster; what they need is a unified knowledge layer they can all trust — one canonical, point-in-time-correct model of the entities that drive trading: grades, curves, cargoes, vessels, locations, counter parties, and assets.

• Design the commercial semantic layer and ontology: canonical entities, relationships, and business definitions shared across desks.
• Establish golden-source master and reference data for commodities (instruments, curves, grades, locations, counter parties, assets) with clear ownership and stewardship.
• Define data products and contracts (schema, SLA, quality, lineage) on top of the enterprise medallion and Gold layer, so consumers get documented, trusted datasets.
• Build point-in-time and bitemporal correctness into the foundation so research and backtests are reproducible and free of lookahead bias.
• Drive entity resolution across messy physical and financial data (the same vessel, cargo, or counter party across sources).
• Set data modeling, quality, and governance standards, and push validation upstream instead of leaving it to strategists.
• Partner with the enterprise data team as the delivery engine, while serving traders, market analysts, and data scientists as the primary consumers, so one knowledge layer feeds both human decisions and ML/GenAI.

• One trusted, governed source of truth for commercial data, so teams stop rebuilding and reconciling
• Traders, analysts and data scientists consume documented data products with known lineage, freshness, and quality
• Research is reproducible: point-in-time correct, auditable, and free of lookahead
• The semantic layer measurably accelerates new desk use cases and the trader interface
• Duplicate knowledge systems retired against a single shared model

• Legally authorized to work in the job posting country
• Bachelor’s degree in Information Technology, Computer Science, Business, Engineering, or a related field
• 7 or more years of experience as a principal-level data architect designing semantic layers, canonical data models, or master/reference data for a data-intensive business
• Advanced proficiency in data modeling, ontologies, and data-product and data-contract design
• Advanced proficiency in point-in-time and bitemporal data and why it matters for trustworthy analytics and research
• Travel up to 5% of the time

• Track record building golden-source reference data and entity resolution across…