The short answer: Data architects decide how an organization's data is modeled, stored, moved and governed, and they get many teams to follow those decisions. The same title covers different jobs. At JPMorgan Chase it is an internal architecture role in Technology. At Microsoft most openings are customer-facing Cloud Solution Architects, and at Deloitte the Data Architect serves clients on a project basis while the Enterprise Data Architect is internal. None of the three publishes a data architect interview loop, and candidate reports for this role are almost nonexistent. What the companies do publish is their general hiring process and their job postings, so this page uses those and says where the few candidate reports fit in. Last reviewed October 2026.
Data Architect interview: quick answers
| Question | Short answer |
|---|---|
| What does a data architect do? | Defines data models from conceptual to physical, sets the target-state data architecture, and builds governance such as lineage, quality, access and retention into designs. |
| What do the interviews test? | No company here publishes a data architect loop. The postings point to data modeling, platform design, governance and behavioral stories. For Microsoft Cloud Solution Architects, candidate reports describe a design round on a customer scenario. |
| Can I use AI tools? | Microsoft says to show your own skills without outside assistance unless explicitly permitted. Deloitte says not to use AI tools unless it tells you otherwise. JPMorgan Chase publishes no rule; ask your recruiter. |
| How much do data architects make? | Posted US base ranges run from $86.1K for a Microsoft Cloud Solution Architect at IC3 to $325K for a JPMorgan Chase Senior Principal Architect. |
| Is it customer-facing? | It depends on the company. JPMorgan Chase's roles are internal. Most Microsoft openings are customer-facing, and Deloitte's Data Architect is client consulting while its Enterprise Data Architect is internal. |
| How is it different from a data engineer? | Data engineers mostly build and run pipelines. The architect postings here focus on models, standards, target states and governance across many teams, though Microsoft's internal architect still writes production code. |
What data architects do
JPMorgan Chase's Lead Data Architects define conceptual, logical and physical models, glossaries, metadata and a semantic layer for domains such as wealth management. Principal architects set systems-of-record strategy and canonical definitions across products, and require an updated model, a named owner and an approved data contract before material schema changes. Its Data Domain Architect titles are a separate data and analytics family.
Microsoft's Cloud Solution Architects lead architecture design sessions, proofs of concept and production deployments for Azure customers on Fabric, Azure Databricks, Cosmos DB and Purview, and their postings include sales-motion duties. The Atlanta Innovation Hub Enterprise Architect is presales. The internal Principal Data Engineering Architect builds Microsoft's marketing data products on Fabric and writes production code.
Deloitte's Data Architect leads a client's domain data architecture: Data as a Product strategy, lakehouse, virtualization and mesh patterns, and migration off legacy ETL such as Ab Initio and Informatica. It is hired under the Project Delivery Talent Model, with employment tied to the project. The internal Enterprise Data Architect defines canonical models and semantic standards across Deloitte's ERP, CRM, MDM and analytics platforms.
How 3 companies run the data architect interview
The short version of each process. Each company guide has the full detail, with links to the company's own pages, and says which parts come from candidate reports.
| Company | Screens | Main interviews | AI tools | What stands out |
|---|---|---|---|---|
| JPMorgan Chase | Assessments only if the role needs them | Several interviewers over multiple rounds; content not published | No published rule | Internal roles; Data Domain Architect is a separate family |
| Microsoft | Application review; brief screen for some roles | Most loops 2 to 4 conversations; CSA design round (candidate reports) | Not without explicit permission | Most openings are customer-facing Cloud Solution Architects |
| Deloitte | Recruiter review of your application | Two to three rounds: behavioral, technical or case-based | Not unless Deloitte explicitly says so | Client Data Architect is hired for a project only |
The interview types in depth
Data modeling
Expect to model a domain at several layers and defend the choices. None of the three companies publishes this round, but every posting here asks for it: JPMorgan Chase asks for conceptual, logical and physical modeling with SCD and event history, Microsoft's internal architect for logical, physical and semantic models, and Deloitte's Enterprise Data Architect for canonical models.
A structure that works: agree the business questions, set the grain, name the entities and their identifiers, then decide what changes over time and how you keep history. Finish with how consumers query it and how the schema evolves without breaking them.
Data platform and architecture design
This is a system design discussion for data: sources, ingestion, storage, serving and governance, with trade-offs. For Microsoft Cloud Solution Architects, a candidate guide and one 2022 London candidate report describe a design round on a customer scenario such as data modernization or migration.
Start with goals and constraints, including data volumes, freshness, compliance and recovery objectives. Sketch the target state, size it with real numbers, and give a migration path from today. Name what you would trade away for cost and what you would not.
Governance and data risk
Architects own the rules around data, so expect questions on access control, lineage, quality, privacy and retention. JPMorgan Chase's postings ask for data risk controls such as tiering, retention, destruction and localization. Deloitte's internal role asks for metadata, lineage and stewardship, and Microsoft's postings name Purview.
Answer with the control, where it lives in the design, who owns it and how you would prove it works to an auditor.
Customer and case scenarios
Customer-facing and consulting roles add a scenario round in some loops: a client asks an open question and you work it through. Deloitte uses case interviews for entry-level Consultative Offerings roles and does not say whether experienced technology hires get one. Microsoft's CSA candidate reports describe case studies.
Clarify the problem, state your assumptions, lay out the options, recommend one, and give next steps and the expected impact. Speak to the client's business as well as the technology.
Behavioral
Every loop has behavioral questions. Deloitte points candidates to the STAR method, JPMorgan Chase asks for measurable results, and Microsoft says interviewers assess respect, integrity, accountability and growth mindset.
Data architect stories that work: a migration you led, a standard you got several teams to adopt without authority, a design an architecture review rejected, and a customer or stakeholder who pushed back.
Pay in US job postings
US salary ranges from each company's own postings for this role, seen October 29, 2026. They are base salary unless the company guide says the posting lists total salary. Each company guide has the postings and links.
| Company | Titles in the postings | Posted US pay range |
|---|---|---|
| JPMorgan Chase | Lead Data Architect, Principal Architect - Data and Senior Principal Architect, Data & Analytics Services | $137.75K to $325K |
| Microsoft | Cloud Solution Architect (data platform), Enterprise Architect - Data Architecture, and Principal Data Engineering Architect | $86.1K to $304.2K |
| Deloitte | Data Architect (Deloitte Consulting, client-facing) and Enterprise Data Architect (Deloitte Global, internal) | $95.6K to $306.8K |
Entry-level paths
- Microsoft: The MCAPS Early in Profession program is a three-year program for future sales and technical leaders; its US page lists no Cloud Solution Architect or data architecture track.
- Deloitte: Entry-level Consultative Offerings candidates go through case interviews; no campus program for data architecture was found.
AI tools in data architect interviews
Rules on AI differ by company, so ask your recruiter before every round.
JPMorgan Chase. JPMorgan Chase does not publish a rule on AI use in interviews; ask your recruiter. Separately, several of its architect postings expect you to use enterprise-authorized AI and agentic tools on the job, with human validation, traceability and auditability.
Microsoft. Microsoft says: "We encourage candidates to use AI tools responsibly and ethically during preparation, as long as it reflects their true capabilities. During assessments and interviews, candidates should demonstrate their own skills without outside assistance unless explicitly permitted."
Deloitte. Deloitte says: "Unless we explicitly tell you otherwise, you should complete your interview by yourself, without seeking assistance from others or using artificial intelligence tools." It lists using AI to research Deloitte, refine resume formatting and generate practice questions as acceptable.
A 6-week prep plan
| When | Focus |
|---|---|
| Week 1 | Read your target company's postings and guide. Note whether the role is internal, customer-facing or client consulting, its stack, its AI rule, and ask your recruiter which rounds to expect. |
| Week 2 | Data modeling: model two domains from conceptual to physical, with identifiers, history (SCD and events) and conformed dimensions. |
| Week 3 | Platform design: warehouse, lake and lakehouse trade-offs, streaming versus batch, and sizing with real numbers. |
| Week 4 | Migration and modernization: plan a move off a legacy warehouse or ETL tool, with waves, parallel runs and a roadmap. |
| Week 5 | Governance: access control, lineage, quality, privacy and retention, and how each shows up in a design review. |
| Week 6 | Behavioral stories and mock sessions: a migration, a standard adopted, a rejected design and a customer pushback, then two practice design sessions out loud. |
Courses for each round
- Data Engineering System Design
- Data Warehousing and Lakehouses
- Data Modeling
- Snowflake
- Streaming Data Engineering
- Apache Spark
Practice questions
Questions from our bank, then practice prompts from the company guides.
Questions from our bank
From the DataInterview question bank, picked for data architect interviews. Open one to answer it.
Security Governance Features
CIO Cloud Migration Guidance
Azure vs AWS Decision
Big Data 5 Vs
Azure Space Data Center
Define Growth Mindset
Database Normalization Steps
Table Joins Differences
Address Feature Gap
Data modeling practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
Model client accounts and their trades for a wealth management data product, from conceptual to physical.
Sample Answer
Start with the conceptual entities: client, account, instrument, trade and position, and how they relate. In the logical model, give each entity a stable identifier and decide which attributes change over time. Keep account attributes as a slowly changing dimension and trades as immutable events, so you can rebuild any past state. In the physical model, pick partitioning by trade date and conformed dimensions shared with other domains. Close with the glossary terms and the semantic layer consumers query.
Model campaigns, touchpoints and leads for a marketing data product on Fabric, from logical to semantic.
Customer is defined differently in the ERP, CRM and MDM systems. Build a canonical model.
Architecture design practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
Design the target-state data architecture for a commercial card and B2B payments business.
Sample Answer
Start with the domains: cards, accounts, transactions, merchants and billing, each owned by a team that publishes data products. Land raw data in a lakehouse on Databricks and AWS, with bronze, silver and gold layers, and serve curated products through tables and APIs. Use streaming for authorizations and batch for statements. Put lineage, quality checks and access control in the platform so each domain does not rebuild them. Finish with the migration path from today's systems and how you measure progress.
A retail customer runs an on-premises data warehouse and wants to modernize on Azure. Run the design session.
A bank client runs hundreds of Informatica batch jobs into an on-premises warehouse and wants a lakehouse. Design the target state and the migration approach.
Governance practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
A new dataset holds client data from several countries. How do you build in localization and retention?
Sample Answer
Classify the data first and assign a tier, since controls follow the tier. Keep data that must stay in a country in storage in that region, and send only permitted fields elsewhere. Set a retention period per dataset with automated destruction, and log what was destroyed. Put these checks in the design review so they are agreed before build, and keep the evidence ready for audit.
A customer wants to know what data it has across its estate and who uses it. How do you set that up?
Set up stewardship for a client's critical data elements.
Behavioral practice
Prompts we wrote for data architect behavioral rounds, with what a strong answer covers.
Tell me about a data platform migration you led.
Sample Answer
Give the scope in numbers, the target state, how you sequenced the move and ran old and new in parallel, a risk that hit and how you handled it, and the measured result.
Describe a data standard you got several teams to adopt.
Tell me about a time an architecture review or a client rejected your design.
Tell me about a time you had to explain a data architecture to non-technical leaders.
Are you ready for data architect interviews?
1 / 5Can you design a data platform for a company (ingestion, storage, serving, governance) with trade-offs?
Company guides
Read the guide for each company you are interviewing with:
- JPMorgan Chase: JPMorgan Chase hires data architects into Technology as Lead Data Architects and Principal Architects. It publishes only a general process, and candidate reports are scarce.
- Microsoft: Most Microsoft data architect roles are customer-facing Cloud Solution Architects. Expect 2 to 4 conversations, and for CSAs a design discussion on a customer data scenario.
- Deloitte: Deloitte's Data Architect is a client-facing consulting role hired for a project, and its Enterprise Data Architect is internal. Expect two to three rounds.
Frequently Asked Questions
What is a data architect interview like?
None of JPMorgan Chase, Microsoft or Deloitte publishes a data architect loop. Their general processes run from 2 to 4 interviews, and the postings point to data modeling, platform design, governance and behavioral questions. Ask your recruiter for the exact rounds.
Are there candidate reports for data architect interviews?
Very few. We found none for JPMorgan Chase or Deloitte data architects. For Microsoft, one candidate guide and one 2022 London Cloud Solution Architect report describe a design or case-study round.
Can I use AI tools in data architect interviews?
Microsoft says candidates should demonstrate their own skills without outside assistance unless explicitly permitted. Deloitte says to complete the interview without AI tools unless it tells you otherwise. JPMorgan Chase publishes no rule, so ask your recruiter.
How much do data architects make at JPMorgan Chase, Microsoft and Deloitte?
In US postings seen in October 2026, JPMorgan Chase lists $137.75K to $325K base from Lead Data Architect to Senior Principal Architect. Microsoft lists $86.1K to $304.2K across Cloud Solution Architect and Principal Data Engineering Architect postings. Deloitte lists $95.6K to $171.3K for Data Architect and $155.6K to $306.8K for a Manager-level solution architect.
Is a Cloud Solution Architect a data architect?
At Microsoft, the data platform Cloud Solution Architect does data architecture for customers: design sessions, proofs of concept and production deployments on Azure. It also carries sales-motion duties, which an internal architect role does not.
What is the difference between a data architect and a data domain architect?
At JPMorgan Chase, Data Architect titles sit in Technology and own models, target states and governance. Data Domain Architect titles sit in Data & Analytics, and the open US postings read as finance data, BI and requirements work. Read each posting to see which you are applying for.
Do I need consulting experience?
Only for the client roles. Deloitte's Data Architect delivers for a client project and Microsoft's Cloud Solution Architect postings prefer 4+ years customer-facing. JPMorgan Chase's architect roles and Deloitte's Enterprise Data Architect are internal.
