Direct Client - Cloud Architect - Purview

Cloud Architect - Direct Client Role
Remote : Travel within USA for 3 client site visits in this 4 months (Expenses paid)
Project phase is for 4 months – Extendable for next 12 months 

Any visa works - C2C -

Long term contract Role if the candidate performs well in initial 4 months 

Senior Cloud Architect - Healthcare Data Governance
1.    Architecture and Strategy
a.    Own the end-to-end target state architecture for workstreams.
b.    Evolve and publish the reference architecture across catalog, lineage, classification, AI accelerators, agent orchestration, observability and consumer surfaces.
c.    Define integration patterns for identity, data access, model endpoints, APIs, vector stores and cross-domain workflow boundaries.
d.    Make build, buy and reuse calls; document Architecture Decision Records (ADRs) with explicit trade offs and risk posture.

2.    Microsoft Purview and Data Governance
a.    Govern the Purview rollout across Snowflake, SharePoint, S3, ADLS, etc. - via supported connectors.
b.    Lead lineage engineering strategy including Atlas API ingestion patterns and lineage stitching.
c.    Drive Critical Data Element identification, sensitivity labelling, DLP policy alignment, and stewardship workflows.
d.    Own the design of consumer surfaces such as NLQ over Snowflake, Institutional Knowledge Search and the Data Contracts framework.

3.    Generative and Agentic AI Platform
a.    Set the architectural direction for the agentic platform: agent framework selection, tool calling, routing, memory, guardrails and reliability patterns.
b.    Lead Microsoft Foundry blueprint validation workshops covering compute, cost, compliance and integration with existing systems.
c.    Define RAG patterns over Purview governed metadata and Snowflake, embedding strategy, vector store choice, evaluation harnesses and prompt or model governance.
d.    Codify HITL checkpoints, observability, RBAC, ethical guardrails and incident response for production grade agentic workflows.
e.    Sponsor reusable accelerators (classification agent, business context extraction, metadata quality monitoring, intelligent search) so each business group does not rebuild the same primitives.

4.    Delivery Leadership
a.    Run weekly working sessions and bi-weekly steering committees with stakeholders.
b.    Unblock the India delivery team (Data Governance Architect, Purview Engineer, Python and Gen AI Engineers, DevOps, QA) on technical and design questions.
 
c.    Set quality gates for releases, evaluation harness adoption and architecture conformance.
d.    Translate emerging client priorities into prioritized work for the monthly capacity envelope of each SOW.
e.    Coordinate transition to managed services or follow on SOWs without scope drift.

5.    Stakeholder and Commercial
a.    Act as the Forward Deployed Engineer Accion Labs in front of customer architecture, security, privacy and procurement stakeholders.
b.    Support account growth by surfacing adjacent opportunities.
c.    Work with the Accion Delivery Manager and Project Manager on margin, capacity and risk governance.

Required Experience
Area    Requirement
Data Platform Dept    Hands on architectural ownership of at least one large scale data catalog, governance or lakehouse program (Microsoft Purview strongly preferred).
AI and Agentic Depth    Production exposure to at least one Gen AI program covering RAG, agent orchestration, evaluation harnesses and HITL controls; Microsoft Foundry or Azure OpenAI strongly preferred.
Cloud and Integration    Architect level experience with Azure (preferred), plus working knowledge of AWS or hybrid topologies; comfort with Snowflake, SharePoint and SaaS integration patterns.
Regulated Industry    Healthcare, life sciences, financial services or similarly regulated environment with HIPAA, GxP, SOC 2 or equivalent control frameworks.
Delivery Methodology    Track record running Agile or Agile hybrid delivery with offshore pods, ARB processes and steering level reporting.

Required Technical Skills
Capability    Specific Technologies and Patterns    Priority
Data Governance and Catalog    Microsoft Purview (Atlas API, scan rules, classifications, lineage, collections, sensitivity labels), Apache Atlas, dbt manifests    Required
Cloud Platforms    Azure (Data Factory, Functions, Key Vault, App Insights, Log Analytics, Foundry, OpenAI), AWS (S3, EC2 for SHIR), hybrid networking    Required
 
Data and Analytics    Snowflake (Cortex, Access History, Query History), SQL, Power BI and Tableau metadata, semantic models    Required
Generative and Agentic AI    Azure OpenAI, Microsoft Foundry, Microsoft Agent 365 SDK, LangChain or LangGraph, LlamaIndex, vector databases, embeddings, RAG patterns, prompt and tool schema design, evaluation harnesses (DeepEval, promptfoo, RAGAS)    Required
Programming and APIs    Python (architecture and code review level), TypeScript and JavaScript (familiarity), REST and async APIs, message queues, Redis    Preferred
DevOps and Platform Engineering    Terraform or Bicep, Azure DevOps or GitHub Actions, Docker, Kubernetes, Azure Container Apps, CI/CD for prompts and models    Preferred
Observability and Reliability    Application Insights, Log Analytics, SLO and SLI design, runbook authoring, ServiceNow integration patterns    Required
Identity and Security    Azure AD, managed identities and service principals, RBAC, network policies, private endpoints, secrets management    Required
Healthcare and Regulatory Patterns    HIPAA aware solution patterns, PHI handling, Business Associate Agreement constraints, sensitivity labelling and DLP    Strongly Preferred

Generative AI and Agentic Platform Expectations
Aspect    Expectation From the Architect
Usage Patterns    Retrieval Augmented Generation over enterprise knowledge bases such as SharePoint and Snowflake, agentic workflows for metadata enrichment and data discovery, and AI native applications anchored on the reusable platform.
Grounding and Data Foundations    Designing solutions that depend on data quality, metadata coverage, lineage, access control and sensitivity labelling. Architect must be fluent in how Purview metadata grounds Gen AI surfaces.
Operating Model and Governance    Establish reusable pattern libraries, prompt and model governance, golden test sets, regression harnesses, drift detection, and a Center of Excellence ways of working with teams.


Interview  Guidelines :
 Please keep a ID handy as a part of HR process to present/show in the interview 
•    It will be a Technical Video Round of interview - video will be recorded. 
•    We request all the candidates not to use headphones and Virtual/Blur Background during Interview
•    Audio and Video Only from one PC/Laptop - avoid dual screens 
•     1st Round with Accion Labs 
•     2nd Round will be with Customer 
•     We cannot tell the Client name we are exclusive vendors its with our Direct Client so please don’t worry on dual submission