Six very different roles share one job title. Here is what each one does, what to look for in a CV, and where organisations get them wrong when they hire.
Data Architect is one of the most misleading job titles, as there are so many types. Some are focused on particular platforms such as Microsoft Fabric. Others have spent their whole careers in analytics and never touched an operational system. Before you can successfully recruit a Data Architect, you must first understand the nuances.
These are the six data architecture roles we see most often, what each one does, and where organisations tend to get them wrong when they hire.
Also seen as: Enterprise Information Architect, Head of Data Architecture
An Enterprise Data Architect sets the data strategy, target state and roadmap for the whole organisation, or for a significant part of a much larger one. They decide how data is organised, owned, shared and governed across the business. More and more, they are also the person accountable for whether the organisation’s data is in any fit state for AI.
You should see data strategies and roadmaps tied to a business outcome, enterprise data models, data domains, and often a data mesh or data product operating model. You should also see evidence of standing in front of executives and securing investment. Their breadth covers operational, analytical and reference data, not just one of them.
A lot of people call themselves Enterprise Data Architects when their whole career has been spent in analytics. If every example in the CV is a warehouse, a lakehouse or a reporting estate, you are looking at a Data Platform Architect with a very generous job title.
Also seen as: Solution Data Architect, Senior Data Modeller
A Data Architect works at project and programme level. They design the conceptual, logical and physical data models for a solution, and they make sure the data works for the business process, not only for the reports that come out of the end of it.
Look for conceptual models agreed with business stakeholders in their own language, and logical and physical modelling in tools such as ERwin, ER/Studio or SqlDBM. Look for relational modelling for OLTP systems, and dimensional or Data Vault modelling for analytics. Data migration and source system integration usually appear somewhere too.
The tell is OLTP. Order management, core banking, policy administration, stock control: operational systems where the data model matters because the business runs on it. If all of the experience is pipelines and warehouses, they are not a solution level Data Architect, however many times the title appears on the CV.
Also seen as: Cloud Data Architect, Lakehouse Architect, formerly Data Warehouse Architect
The Data Platform Architect designs the platform that ingests, stores, processes and serves data across the organisation, now almost always in the cloud. This is where most Data Warehouse Architects have gone. The warehouse became the lakehouse, and they moved with it.
You will see Databricks, Snowflake, Microsoft Fabric, Google BigQuery, Azure Synapse or AWS Redshift. You will see lakehouse and medallion architectures, Delta Lake or Apache Iceberg, ingestion and transformation with Azure Data Factory, dbt or Spark, and Kafka where streaming is involved. The better ones will also talk about cost, security and scale, because those are the conversations that decide whether a platform survives its second year.
The architect designs the platform, and the engineer builds pipelines on it. Plenty of senior Data Engineers now carry the architect title without ever having made the design decisions. Ask them why the platform looks the way it does, not what they built on it.
Also seen as: BI Architect, Analytics Architect, Reporting Architect
The Analytics and BI Architect designs how the business actually consumes its data. That covers semantic models, reporting, self service analytics, and the standards that stop finance and operations turning up to the same meeting with different numbers.
Look for Power BI, Tableau, Looker or Qlik, shared semantic layers, self service strategy, and report governance at scale. You should also see close working with finance and operational teams on metrics and KPIs.
The BI Architect works at the consumption end, and the Platform Architect works on everything underneath it. Some people do both well. Most are firmly one or the other, and the CV will show you which.
Also seen as: Principal DBA, Database Engineer
The Database Architect is very technical and works at the physical layer. The focus is performance, availability, resilience and migration, increasingly on cloud managed databases rather than servers in a basement.
Expect real depth in one or two technologies rather than a long list: Oracle, SQL Server, PostgreSQL, Azure SQL, Amazon Aurora, Cosmos DB, MongoDB or DynamoDB. Around that you will see performance tuning, high availability, replication, backup and recovery, and at least one large migration, usually from on premises to cloud.
A Database Architect designs how the database performs. A Data Architect designs what the data means. Very few CVs genuinely show both, and the ones that claim to usually show neither in much depth.
Also seen as: Data Governance Lead, Data Quality Architect, Metadata Architect
The Data Governance and Quality Architect designs the frameworks, tooling and processes that make data trustworthy. That means ownership, lineage, cataloguing, quality rules, master and reference data, and regulatory compliance. It has gone from a nice-to-have to a necessity since organisations started pointing AI at their data and discovering what was in it.
Look for Microsoft Purview, Collibra, Alation, Informatica or Databricks Unity Catalog. You should also see data ownership and stewardship models, master data management, and data quality measurement and remediation, often shaped by GDPR, BCBS 239 or a sector regulator.
If the CV is mostly policy documents and steering committees, you have found a very good governance manager. You have not found an architect.
If you are hiring for a data role and are not sure which of these you actually need, please get in touch. It is surprisingly common for the job description to describe one of these roles while the job title asks for another.
Ben Clark
bclark@konvergent.co.uk
