- Adviesdiensten
DWH migration to cloud - 5-Month implementation
Organizations around the world are rapidly moving to the cloud, since cloud lets you store, analyse, and share data in ways that were, up until a few years ago almost unthinkable.
Like many organizations, you’re facing challenges with your existing data warehouse solution and are looking for alternatives. You spend days trying to: • get data to end-users • your data is scattered in data silos, • your on-premise infrastructure is outdated and offers limited scalability and elasticity • Lack of resources means you can’t address new business cases without investment in new HW and SW. • Your IT is struggling with ongoing maintenance and optimizations to keep the lights on...
Your leadership team wants faster insight into data and you are overwhelmed with new requests (resource contention). Microsoft Azure Synapse analytics offers a new way of thinking and opens up huge opportunities as you move from on-premises to the cloud.
This offer provides you with a guided migration process of your on-premise data warehouse to the modern cloud platform Microsoft Azure Synapse analytics. Our common goal is to help you address the overall migration approach, build a modern cloud data warehouse and answer the questions of overall architecture, project management, services, and infrastructure cost.
Outcomes During the delivery process, we will define the project scope and success criteria. The idea is to define End2End architecture including modern ETL/ELT tools for data ingestion and transformation to take the cloud advantages and test your existing or new BI/data science tools against cloud infrastructure.
Implementation phase (3 main phases) • Design (hint: spend enough time on this phase) and detailed business scope • Detailed target architecture: • Overall data platform landscape and flow of data in detail • Data ingestion/transformation • Reporting, data science platforms and 3rd party integrations • The exact definition of the data flow • Setup cloud platform and security • Detailed project plan and implementation • Development • Continuous data feeds • Initial load • Reconciliation • UAT and Parallel Run