Data architect is the senior technical role responsible for the overall design of an organisation's data ecosystem — the schemas, storage layers, data flows, and governance frameworks that allow data to be collected reliably, stored efficiently, and used effectively across analytics, ML, and product functions. The role sits above data engineering in the technical hierarchy and focuses on the strategic and structural decisions that constrain how data engineering work is done.
What the work actually splits into
Enterprise data modelling. You design the conceptual, logical, and physical data models that define how business entities and their relationships are represented across the data estate. Decisions made here ripple through every downstream system — poorly designed data models create technical debt that persists for years.
Data platform architecture. You design the overall technical architecture of the data platform — the lakehouse or warehouse layer, the ingestion and transformation pipeline architecture, the serving layer for analytics and ML, and the integration patterns that connect operational systems to the data platform. Modern stacks (Databricks, Snowflake, BigQuery + dbt + Airflow) need coherent design rather than accumulation.
Data governance and standards. You establish the technical standards that engineering teams follow — naming conventions, data quality expectations, schema evolution policies, access control patterns, and the data contract specifications that define what producers commit to publishing. The architect sets the rules; governance engineering enforces them.
Reference architecture and pattern library. You create reusable architecture patterns that engineering teams can apply without reinventing solutions to problems you've already solved — standard patterns for slowly changing dimensions, event sourcing, feature store integration, real-time analytics pipelines, and so on.
Technical roadmap and vendor evaluation. You evaluate new data technologies and make the architectural decisions that shape the data stack over a multi-year horizon — when to migrate from a legacy warehouse, whether to adopt a lakehouse pattern, which streaming technology fits the organisation's requirements. These are high-stakes, high-cost decisions that require both technical depth and commercial judgment.
The employer landscape
Large enterprise companies across financial services, healthcare, retail, and manufacturing hire data architects to bring order to complex, multi-decade data estates that have accumulated through acquisitions, organic growth, and siloed IT investment. These roles often involve migration from legacy on-premise systems to cloud-native architectures.
Data-intensive technology companies (analytics platforms, SaaS with data products, marketplace businesses) hire data architects to design the foundation that their data products are built on. The scale of these roles is often higher, the stack is more modern, and the pace of architectural change is faster.
Consulting and advisory firms staff data architects on client transformation engagements — helping organisations design target-state architectures, evaluate vendors, and plan migration programmes. High variety; less depth per client.
Financial services and fintech companies hire data architects with specific experience in regulatory data requirements — Basel III data lineage obligations, BCBS 239 risk data aggregation, MiFID II trade reporting — where the architecture has direct regulatory consequence.
What skills actually differentiate candidates
Depth in modern data stack patterns. The best data architects have built production systems on Snowflake, BigQuery, or Databricks, used dbt for transformation, and designed Kafka or Spark-based streaming architectures. Architectural authority without hands-on depth produces elegant diagrams that fail in implementation.
Data modelling rigour. Dimensional modelling (Kimball), data vault, and OBT (one big table) patterns are not interchangeable — the right choice depends on the organisation's query patterns, update frequency, and the skill level of downstream consumers. Architects who can make and justify these choices with evidence are more valuable than those who default to a single pattern.
Technology-agnostic judgment. The data tooling landscape is crowded and fast-moving. Architects who have genuine opinions about when dbt is the wrong choice, or when a data warehouse doesn't need a lakehouse layer, or when real-time is operationally expensive for marginal analytical value — are more trusted than those who endorse the latest stack uncritically.
Communication with non-technical stakeholders. Data architecture decisions affect business teams, finance, legal, and compliance. Architects who can explain the business implications of technical choices — and translate regulatory or business requirements into architectural specifications — are significantly more effective in enterprise contexts.
Five things worth checking before you apply
Is this a greenfield or brownfield architecture? Designing a data platform from scratch is different from rationalising a legacy estate with dozens of source systems and years of accumulated technical debt. Both are legitimate; know which one you're walking into.
What is the relationship with data engineering? Data architects who lack authority over the engineering work implementing their designs are in an advisory role, not an architectural one. Understand whether architectural standards are enforced or merely recommended.
What is the primary use case driving this role? Analytics and BI, ML and data science, operational data products, and regulatory reporting each imply different architectural priorities. Roles without a clear primary driver often produce architectures optimised for the wrong thing.
What is the cloud platform and existing data stack? The specific stack tells you what migration debt exists and what technical skills will be immediately applicable. A Snowflake-on-AWS architecture requires different expertise than a Databricks-on-Azure one.
Is there a data governance or stewardship function? Data architects who also own governance implementation have a much larger scope than those who are purely technical. Know whether governance is in scope.
The bottleneck at each level
Junior data architects (often titled senior data engineer making the transition) are bottlenecked by pattern breadth. They design well for problems they've seen before but struggle with novel combinations of requirements that don't map neatly to familiar patterns. Breadth of hands-on experience with different architectures is the fastest path through this bottleneck.
Mid-level data architects are bottlenecked by organisational influence. The most technically correct architecture is worthless if the engineering teams don't implement it. Getting engineering teams to follow architectural standards — especially standards that create short-term friction — requires credibility, relationship, and patience.
Senior data architects are bottlenecked by strategic horizon. At this level, the work is less about specific technology choices and more about the multi-year data strategy — what capabilities does the organisation need in three years, what architectural decisions today enable or constrain those capabilities, and how do you sequence investment to get there. This is more executive stakeholder work than technical design.
Pay and level expectations
Remote data architect salaries in the US range from $150,000–$190,000 at mid-level to $190,000–$250,000 at principal/staff architect level. Financial services and healthcare roles with regulatory consequence tend to pay toward the upper end. The function is specialised enough that strong candidates have significant negotiating leverage.
European remote roles typically pay €90,000–€135,000 depending on seniority, industry, and the maturity of the data function.
What the hiring process looks like
Data architect interviews typically include a system design exercise (design a data platform for this scenario), a data modelling exercise (here is a business domain — model it for an analytics use case and explain your choices), and a technology evaluation discussion (compare these two approaches for this requirement — what would you choose and why?).
Senior roles often include a presentation of past architecture work — a system you designed, the trade-offs you navigated, and what you'd do differently. References from engineering teams who have worked within architectures you've designed carry significant weight.
Red flags and green flags
Red flags: Architecture is purely theoretical — no implementation authority and no feedback loop from engineering on what actually works. Stack is heavily vendor-driven with significant lock-in and no data migration strategy. Data estate has grown entirely organically with no governance, and there's no leadership appetite to enforce standards.
Green flags: Engineering leadership that actively participates in architectural decisions and holds teams accountable to architectural standards. Existing data platform with clear pain points that an architect would have authority to address. Cross-functional mandate involving both technical and business stakeholders. Clear data strategy at the executive level.
Gateway to current listings
Use the listings below to find current remote data architect openings. Titles vary significantly — "data architect," "principal data engineer," "staff data engineer," "enterprise data architect," and "chief data architect" can all describe similar roles depending on company size and naming convention. Read for scope — are you designing systems or advising on systems? — rather than anchoring on the title.
Frequently asked questions
What is the difference between data architect and data engineer? Data engineers implement the pipelines and infrastructure that move and transform data. Data architects design the overall system that data engineers build within — the schemas, the platform topology, the standards and patterns. The relationship is similar to software architect versus software engineer.
Do data architects write code? Yes, though less than data engineers. Proof-of-concept implementations, reference architecture validation, and data model prototyping all require hands-on technical work. Architects who can't write working code struggle to design architectures that are practical to implement.
Is data architecture a growing field? Yes — the proliferation of data sources, the rise of ML and AI applications, and increasing regulatory requirements around data quality and lineage have all driven demand for senior data architectural thinking. The role is less commoditised than data engineering because the judgment required is harder to automate.
What certifications are useful for data architects? Cloud provider data specialisation certifications (AWS Data Analytics, Google Professional Data Engineer, Azure Data Engineer) are useful signals. Vendor-specific certifications (Databricks, Snowflake SnowPro) are recognised in the market. None of these substitute for demonstrable hands-on experience with production-scale systems.
Related resources
- Remote data engineer jobs — implementation counterpart to data architecture
- Remote data platform engineer jobs — platform infrastructure for the data estate
- Remote cloud architect jobs — cloud infrastructure architecture counterpart
- Remote data governance engineer jobs — governance and quality engineering adjacent role