Article originally written by Scott Emigh. Find him on LinkedIn here.

In this final post on overcoming the major challenges to becoming a great AI partner, we’ll address how the condition of a customers’ data estate can be a serious blocker to AI, and why you should always modernize your customer’s data estate when migrating to cloud. 

Is the state of your customer’s data environment inhibiting or preventing them from cost-effectively implementing AI scenarios? Do they fully understand this situation? One of the biggest challenges to customers adopting cloud productivity solutions such as Microsoft Office 365 and Microsoft Dynamics 365 is the messy state of their identity management environment. Their data estate can cause the same problems for them when it comes to AI.

Incomplete or messy data is just one problem with a customer’s data estate. Other issues such as incompatible data models, overly rationalized transactional DBs, rigid VM-based architectures, and non-performant repositories can all contribute to a data environment that’s unfriendly to AI projects.

I encourage our partners to always think about cloud migrations as an opportunity to modernize the customer’s data estate—not just for transactional performance and economy, but also to enable data warehouse and analytics scenarios. For example, does log data need to migrate from an on premises relational database to Microsoft Azure SQL database, or would it be more cost-effective and better primed for analytics processing if it was migrated to a No SQL environment like DocumentDB?

With Microsoft SQL Server 2008 & 2008 R2 end of support coming July 2019 and Microsoft Windows Server 2008 & 2008 R2 end of support coming January 2020, there’s never been a better time to upgrade, modernize, and transform to current versions of SQL Server, Windows Server, and Azure services. Learn more about the end of support migration program and support here.

Chris Mitchell, US OCP Cloud Solution Architect on our Data & AI team, had this advice for partners approaching a cloud migration project:

  • Are your databases ready for PaaS? The easiest migration to the cloud is by migrating Virtual Machines, but it’s rarely the most cost-effective approach. By migrating databases like SQL Server, MySQL, and Postgres to our Azure database service, you can dramatically reduce overall TCO by leveraging the elasticity and manageability benefits of these services—including out-of-the-box high availability, built-in backup, and simple disaster recovery.
  • Are there any latent scale and performance needs? Oftentimes, some of the existing data solutions have pre-existing performance and scalability issues that won’t be eased by cloud migration. It’s important to understand and consider these existing scale issues to consider alternative technologies. For example, would your existing data warehouse benefit more from a migration to Azure SQL Data Warehouse, or from being redesigned to leverage Azure Data Lake as a component of the overall architecture?
  • Is it time to re-evaluate your multi-tenant design? Similar to analytics scale and performance issues, many partners are migrating solutions that have multi-tenant capabilities. If your tenancy design is using a single database, cloud migration is a great opportunity to break up this design and spread tenants across multiple databases, allowing you to take advantage of features like resource pooling with Azure SQL Database Elastic Pools, which protects you from noisy neighbor scenarios and improves overall scalability.
  • Do any of the systems benefit from architectural innovation? Many of our ISV partners are redesigning existing solutions as they move to the cloud, and use cloud migration as an opportunity to redesign their data architecture in order to leverage technology like Azure Cosmos DB and deliver the agility of a NoSQL platform, in addition to Cosmos DB’s global footprint and limitless scale.

One of the other challenges of successfully implementing AI solutions is gathering the right AI and data science talent. Finding the talent with formal statistical and data science training is becoming increasingly difficult and expensive. Microsoft is investing heavily in readiness and training programs to build your practice’s AI talent. Start with these online Data Science and AI edX courses through Microsoft Professional Program:

If you’re ready to dive into AI, check out our other resources:

  • LearnAI helps you find data science-related content offered directly from Microsoft, or through our vetted list of AI training partners
  • Get certified via the Data Management & Analytics MCSE path. The Machine Learning MCSA is a great way to catalyze readiness and advertise your team’s expertise
  • Download a copy of our Data & Analytics Partner Practice Playbook

I hope you found these tips useful, and remember that Microsoft One Commercial Partner is here to help with your AI practice development efforts.


Learn how we can accelerate your business. Visit the Microsoft NZ Partner Hub at