Topic 5 of 5
Roles and responsibilities for data workloads
Managing, moving, and analyzing data is done by people with different profiles. The exam describes a task and asks which role does it, mainly among database administrator, data engineer, and data analyst. The official training now adds a fourth role: the AI engineer.
Contents
Data roles
Some roles are business-oriented, some are more engineering, some focus on research, and some are hybrids. Each organization may define or name them differently, but these are the most common:
Database administrator
Manages databases: assigns permissions, stores backups, and restores data after a failure.
Data engineer
Manages infrastructure and processes for data integration: cleaning, governance rules, and pipelines to move and transform data.
Data analyst
Explores and analyzes data to create visualizations and charts that support decisions.
AI engineer
Builds and integrates AI-powered features and workflows, working with language models, machine learning pipelines, and data sources.
One person, several roles
In some organizations the same person plays several parts: as an administrator they provision a transactional database, and as a data engineer they build a pipeline that moves that data to a data warehouse for analysis. Other roles exist that aren't covered here, such as data scientist or data architect.
Database administrator
Responsible for the design, implementation, maintenance, and operation of on-premises and cloud database systems.
- Overall availability, consistent performance, and optimization of databases.
- Policies, tools, and processes for backup and recovery after natural disasters or human error.
- Data security: grants privileges and allows or denies user access.
- Increasingly uses AI tools to troubleshoot performance and draft queries in natural language, without replacing their judgment.
Data engineer
Works with stakeholders to design and implement data workloads: ingestion pipelines, cleansing and transformation activities, and data stores for analytics.
- Uses many technologies: relational and non-relational databases, file stores, and data streams.
- Ensures data privacy is maintained in the cloud and between on-premises and cloud.
- Manages and monitors pipelines so data loads perform as expected.
- AI helps generate transformation code and suggest configurations; architecture and data quality decisions remain theirs.
Data analyst
Helps the business maximize the value of its data.
- Explores data to identify trends and relationships.
- Designs and builds analytical models.
- Enables advanced analytics through reports and visualizations.
- Turns raw data into useful insights based on business requirements.
- AI helps summarize reports, suggest visualizations, and generate analytical expressions.
AI engineer
Builds and integrates AI-powered features into applications and data workflows. Works with large language models (LLMs, AI systems trained on vast amounts of text that understand and generate language), machine learning pipelines, and data sources.
- Typical scenarios: chat over company data, content generation, and automated classification.
- Works with data engineers to access and prepare data, and with analysts to surface AI results in reports and applications.
- Uses Microsoft Foundry, Azure's PaaS platform to build, test, and deploy AI solutions.
- The design, evaluation, and responsible deployment of AI remain human responsibilities.
Comparing responsibilities
| Role | Focuses on | Example task |
|---|---|---|
| Database administrator | Availability, performance, security, backups | Grant permissions with GRANT, restore a backup |
| Data engineer | Integration, pipelines, analytical stores | ETL pipeline from the transactional database to the warehouse |
| Data analyst | Models, reports, visualization | Interactive sales report in Power BI |
| AI engineer | AI features over data | Chat assistant over internal documents |
Which services each role uses
Azure SQL and open-source databases: Administrators provision and manage them for transactional apps; engineers use them as pipeline sources; analysts may query them directly.
Azure Cosmos DB: Managed by an administrator or, often, by the developers themselves as part of the app; engineers integrate it into analytics solutions.
Azure Storage and Data Lake: Engineers host data lakes in Blob Storage with a hierarchical namespace.
Azure Data Factory and Microsoft Fabric: Engineers build ETL pipelines and unified analytics solutions on OneLake.
Power BI: Analysts connect to data, build reports and dashboards, and share them.
Azure Databricks and Stream Analytics: Engineers process large volumes with Spark or streaming data; analysts can query from notebooks.
Microsoft Purview: Engineers enforce data governance and track lineage across sources.
Microsoft Foundry: AI engineers build, test, and deploy intelligent solutions.
Exam scenarios
| Task | Role | Why |
|---|---|---|
| Assign permissions to users and restore a database after a failure | Database administrator | Security, backups, and recovery. |
| Build a pipeline that cleans data and moves it to a data warehouse | Data engineer | Ingestion, transformation, and pipelines. |
| Create a dashboard with sales trends for leadership | Data analyst | Exploration, models, and visualization. |
| Add a chat that answers questions about company documents | AI engineer | Integrates language models with data. |
| Plan database recovery after a natural disaster | Database administrator | Backup and recovery plans. |
| Make sure nightly data loads complete successfully | Data engineer | Pipeline management and monitoring. |
| Design an analytical model so the business can build its reports | Data analyst | Analytical models and reports. |
Common exam traps
❓ Who handles database backups and recovery?
The database administrator.
❓ Who monitors data pipelines?
The data engineer, who also designs and builds them.
❓ Does the data analyst build the pipelines that feed the data warehouse?
No. That's the data engineer; the analyst explores, models, and visualizes.
❓ Who ensures data privacy between on-premises and the cloud?
The data engineer, according to the official role description.
❓ Which platform does the AI engineer use to build and deploy solutions?
Microsoft Foundry.
❓ Can one person hold several roles?
Yes. In some organizations the administrator also acts as a data engineer.
Did you get this topic?
Test what you just learned
A person builds a process that copies orders from Azure SQL Database every night, cleans them, and loads them into an analytical store. Which role are they performing?
Ready to test "Data roles and responsibilities"?
Original questions, each with a full explanation.
Start a mock exam