Cloud & Data Platform Architect
- Diverse Team
8 Hours Ago
NA
W2, C2C
San Ramon-CA
5-10 Years
Required Skills: Databricks, Snowflake, Kafka,
Job Description
We run production cloud and data platforms for enterprise clients — multi-region GCP estates, BigQuery warehouses feeding hundreds of downstream consumers, and BI layers that executives look at every morning. Those platforms are increasingly not single-cloud. Workloads, identity, and data land in both GCP and Azure, and someone has to own the shape of the whole thing rather than one slice of it.
This is that owner. It is a hands-on architect role, not a slideware role. You will design the landing zones, review the Terraform, sit in the incident call when the warehouse queues, and then present the remediation to a client CIO the same week.
What you will own
Platform architecture
• Target-state architecture for cloud infrastructure and data platforms across GCP (primary) and Azure (secondary but real)
• Landing zone design: org/subscription hierarchy, network topology, connectivity (Interconnect / ExpressRoute), DNS, egress control
• Identity and access architecture spanning Google Cloud IAM and Microsoft Entra ID, including federation and workload identity
• Cost architecture — capacity vs on-demand models, BigQuery editions and slot reservations, Azure reservations, chargeback/showback
Data platform
• Warehouse and lakehouse architecture: BigQuery, ADLS Gen2, Synapse or Fabric, Databricks where relevant
• Ingestion and orchestration design: Pub/Sub, Dataflow, Datastream, Cloud Composer / Airflow, Azure Data Factory
• Transformation layer standards: Dataform and/or dbt — repo structure, environments, promotion path, CI gates, testing
• Data governance: catalog, lineage, classification, retention, PII handling, and the controls that make a GDPR/DPDP audit uneventful BI and semantic layer
• Looker platform ownership: LookML model architecture, Explore design, PDT/aggregate-awareness strategy, performance tuning
• Access architecture in Looker — model sets, user attributes, row-level security patterns that survive an audit
• Instance administration, upgrades, embed and API usage, and migration work off legacy BI tools
Engineering discipline
• Infrastructure as code as the only path to production — Terraform modules, state strategy, policy-as-code
• CI/CD across Cloud Build / GitHub Actions / Azure DevOps
• Observability and SLOs for both infrastructure and data (freshness, volume,schema, cost anomalies)
• Reliability practice: runbooks, RCA ownership, postmortems that change the design
Client and commercial
• Translate a business problem into an architecture and a defensible estimate
• Lead technical discovery, write solution sections of proposals, and present to client architecture review boards
• Mentor 5–10 engineers; set the technical bar in design reviews and code reviews
Must have
• GCP, deep: BigQuery, GKE, Cloud Run, Cloud Storage, VPC design and VPC Service Controls, IAM and org policies, Cloud Composer, Pub/Sub, Cloud Build, Artifact Registry, Cloud Logging/Monitoring
• Azure, working depth: landing zones per Cloud Adoption Framework, AKS, Azure Storage/ADLS Gen2, Azure Data Factory, Entra ID, Azure Policy, Azure
Monitor, Azure DevOps
• Infrastructure: networking that you can whiteboard from memory — routing, peering, private endpoints, hybrid connectivity, firewall and egress design
• SQL at an expert level, plus Python for tooling and automation
• Terraform in production, with modules you have authored and maintained
• Data modelling — dimensional and denormalised, and the judgement to know when each is right
• Demonstrated ownership of a production platform at meaningful scale: TB-to-PB warehouse, 100+ source systems, or 500+ BI users
• Experience running a multi-cloud or hybrid estate in reality, not just in a diagram
• Client-facing communication strong enough to hold a room of skeptical senior stakeholders
Strongly preferred
• Looker / LookML hands-on: development and administration. This is the single biggest differentiator among otherwise equal candidates
• Dataform or dbt at scale, with a real promotion process between environments
• FinOps track record — a specific cost reduction you can quantify and explain
• Migration experience: on-prem to cloud, or BI tool consolidation (Domo, Tableau, Power BI, Qlik → Looker)
• Exposure to production GenAI or agentic workloads on cloud infrastructure
• Certifications: GCP Professional Cloud Architect or Professional Data Engineer; Azure Solutions Architect Expert (AZ-305); Looker LookML Developer
Nice to have
• Kafka or Confluent; streaming-first architectures
• Databricks, Snowflake, or Fabric comparative experience
• Regulated-industry background (BFSI, healthcare, telco) with the compliance posture that comes with it
• Pre-sales or partner co-sell experience with a hyperscaler field team
