Azure Data Architect

FULL REQUIREMENTS
• 1/Azure Data Architecture and AI-ready data architecture
o The candidate must demonstrate and describe his duration experience on which
project and in what context, for how long duration with Azure Data
Architecture with experience designing the data layer supporting
GenAI/RAG/Agentic applications.
o They are not looking for someone who needs to develop RAG solutions or
AI agents hands-on.
o However, the Data Architect must have experience designing the data
architecture around applications using RAG and/or Agentic AI patterns
o For example, how enterprise data and knowledge are ingested, prepared, stored,
governed and made available for retrieval and AI consumption.
o We are looking for someone who combines strong Azure Data Architecture
with real project experience designing the data layer that supports
GenAI/RAG/Agentic applications
o The candidate must demonstrate his experience with AI-ready data
architecture.
o The candidate must demonstrate his strong experience with Azure Data
Platform architecture ownership and delivery experience.
o The candidate must demonstrate his concrete examples of inheriting a
problematic legacy data platform, separating tactical stabilization from
long-term architecture, implementing incremental processing, working
with genuinely unstructured/file-based data, and evaluating technologies
based on scalability, cost, workload and business requirements rather than
preference with clear examples where architectural decisions were actually
implemented by engineering teams.
o The candidate must demonstrate his relevant experience with designing and
implementing complex data architectures, particularly involving unstructured
data and different file formats
o The candidate must demonstrate his relevant experience with Taking over an
existing problematic data platform and improving
• 2/Designing the data layer supporting GenAI/RAG/Agentic application
• The candidate must demonstrate and describe his duration experience on which
project and in what context, for how long duration with designing the data layer
supporting GenAI/RAG/Agentic applications, production RAG/Agentic data
architecture.
• The candidate must demonstrate his Future-oriented architecture thinking
around AI and natural-language querying of enterprise data
• He must be able to explain his various projects with designing the production data
architecture behind a RAG/Agentic application
• He must be ale to explain his experience designing the data architecture around
applications using RAG and/or Agentic AI patterns.
For example, how enterprise data and knowledge are ingested, prepared, stored,
governed and made available for retrieval and AI consumption.
o The candidate must show and explain his real project where designed or
contributed to the data architecture supporting a RAG, GenAI or Agentic
AI application
o The candidate must explain his project with data architecture supporting a
RAG, GenAI or Agentic AI application :
▪ the project and approximate dates;
▪ what the RAG/GenAI/Agentic application was designed to do;
▪ what The candidate personally designed or owned;
▪ how the source data or enterprise knowledge was prepared for theAI applic
ation;
▪ whether The candidate worked with documents, OCR, textextractio
n, metadata, chunking, embeddings, vector search/retrievalor simila
r patterns;
▪ which storage/search technologies were used;
▪ whether Cosmos DB was part of the AI/LLM architecture, and how;
▪ whether the solution was production-based or a POC.
• 3/Azur
e
Cosmos DBo The can
didate must demonstrate and describe his duration experience on whichproject a
nd in what context, for how long duration with Azure Cosmos DB.Azure Cos
mos DB is fundamental and essential for this role.o The can
didate must be able to demonstrate his experience working with filesand Cosmo
s DB, an essential skill for the data layer.o The can
didate must be able to demonstrate his experience with Cosmos DBhow long
he worked with it, and whether the described decisions arounddata mode
ling, partitioning, indexing, RU/scalability and cost optimizationwere pers
onally .o The can
didate must demonstrate his evidence of the architecture ownershipand depth
required for this role. demonstrated Azure Cosmos DBexperienc
e,o The can
didate must demonstrate his practical experience with Cosmos DB,including
scaling, partitioning, RU optimisation etc.o Cosmos
DB: must be demonstrated hands-on experience deeper knowledgearound sc
alability, throughput/cost optimization and advanced design decisionswas less
convincing.o Azure C
osmos DB : practical topics such as data modeling, partition-keystrategy,
indexing, RU/throughput and cost, scaling, consistency and ChangeFeed.o Mo
re im
portantly, The candidate should be ready to explain when CosmosDB is the
right architectural choice and when it is not.
• 4/Azur
e
Synapse Analytics .o The can
didate must demonstrate and describe his duration experience on whichproject a
nd in what context, for how long duration with Azure SynapseAnalytics
.o The can
didate must demonstrate his projects and expériences with Synapsedepth, sc
alable/configuration-driven ingestion.o The can
didate must demonstrate his experience with Azure-native services,particula
rly Synapse Analytics and the different workload patterns acrossPipelines
, Spark, Serverless SQL, and Dedicated SQL Pools, particularlywith Syna
pse depth, scalable/configuration-driven ingestion.o Synapse
must be demonstrated depth across the different workload patterns andarchitect
ural choices required for this roleo Azure S
ynapse depth – The candidate should be ready to explain whereSynapse f
its, when to use different Synapse capabilities, and how itcompares
with Azure SQL, ADF, ADLS, Databricks/Fabric, etc.
• 5/Azur
e
Databrickso The can
didate must demonstrate and describe his duration experience on whichproject a
nd in what context, for how long duration with Azure Databricks.o The can
didate must demonstrate his experience with Data Engineering, DataWarehousi
ng, Databricks, and Azure platform experience.o The can
didate must describe Azure-native alternatives when Databricks wasremoved f
rom the architecture discussion, with dbt being presented as anETL repla
cement rather than a transformation framework.o Technol
ogy-neutral decisions – they may deliberately remove Databricksfrom a sc
enario and ask how he would solve it using other Azure technologies.o Technol
ogy neutrality: the client is specifically trying to determine whethertechnolog
ies such as Databricks, Synapse or ADF are actually necessary fortheir env
ironment or whether they would introduce unnecessary complexity. Forthe role,
o We need
someone who is comfortable concluding that Databricks may ormay not b
e the right solution if the requirements, scale, cost or existingarchitect
ure do not justify ito The can
didate must demonstrate his relevant experience with Azure data-platform
experience, while remaining technology-neutral when decidingbetween A
zure SQL, Synapse, ADF, Databricks, Lakehouse, etc.;o The can
didate must demonstrate his Abilities to assess when technologiessuch as A
zure SQL, ADF, Synapse or Databricks/lakehouse areappropria
te for the actual data size and use case.
• 6/Azur
e
SQL MI / SQL Server / T-SQLo The can
didate must demonstrate and describe his duration experience on whichproject a
nd in what context, for how long duration with Azure SQL MI / SQLServer /
T-SQL.
• 7/Azur
e
Data Lake / Lakehouse architectureo The can
didate must demonstrate and describe his duration experience on whichproject a
nd in what context, for how long duration with Azure Data Lake /Lakehouse
architecture.
• 8/Azur
e
Functionso The can
didate must demonstrate and describe his duration experience on whichproject a
nd in what context, for how long duration with Azure Functions.
• 9/Stro
n
g architecture ownership,o The can
didate must demonstrate and describe his duration experience on whichproject a
nd in what context, for how long duration with Strong architectureownership
, data modeling, ingestion and platform modernization.o The can
didate must demonstrate his relevant experience with practical datamodeling
and performance/scalability experience.o The can
didate must demonstrate his Strong data-modeling knowledge,including
fact tables, dimensional/star schemas and optimization ofanalytica
l models.o The can
didate must demonstrate his experience producing detailedarchitect
ure documentation, diagrams and technical recommendations.o The can
didate must demonstrate his relevant experience with Tactical andtarget ar
chitecture ownership.o Tactica
l vs. strategic architecture we need someone who can very quicklyseparate
immediate fixes from larger long-term architectural changes andprogress
both in parallel.o The can
didate must demonstrate his Abilities to define both tacticalarchitect

Work arrangement
Yes