Today we are visiting with SpringML’s Sr. Vice President of Strategic Accounts, Piyush Malik. He shares why he is so excited about AI/ML in 2020 and best practices around moving data to the cloud and an example of how the GCP can unlock powerful insights and enable predictive analytics as soon as data is consolidated in the data warehouse.
Megan – Today, I’m delighted to have Piyush Malik, our senior vice president of strategic accounts heading up our Google Practice. I understand that you’ve been working to help our customers bring more data from their data warehouse into the cloud. What are you seeing in 2020 now that customers can maximize the art of the possible and shift their core infrastructure into the cloud?
Piyush – Glad to be here and delighted to be talking to our listeners today. Large amounts of workloads are moving from on-prem to the cloud, integrating legacy data sources and integrating hybrid and multi-cloud patterns into the core infrastructure. Let me outline a few things that stand out in our practice. One of the things that remains foundational is to have a solid business case for whatever we are building.
Data Warehouses are used for managing the business and providing insights in a timely fashion to the end-user. Cloud provides that seamlessly and with less friction and lower start-up costs. In the past, we used to wait for the hardware to be procured, software to be configured, the infrastructure to be built, and lastly reports and analytics to be designed. Even though we can take care of all the infrastructure very quickly. When moving to the cloud, having a clear understanding the business objective is a foundation principle. We need to define what analytics are needed, and we need to involve the stakeholders throughout the process.
Megan – What I hear you saying is that the speed factor is significant, and with that sort of big rock out of the way, the art of the possible can be unlocked. Once you move in your data warehouse into the cloud and you’ve got a lot of data to analyze. What are some of the best practices you can share with our more technical audience who are still struggling with that very first step of moving data to the cloud?
Piyush – Once the business purpose for the warehouse has been established, then comes the selection of the right Cloud Platform. At SpringML, we’ve seen a rapid adoption of Google Cloud. After deciding on the platform, you also need to make your stakeholders aware of what would be the total cost of ownership? It’s not only for the set-up but also for the total cost of ownership in terms of deployment, development, management cost. Next comes simplifying the application deployment process to accelerate the realization of value from data. Integrating structured data along with unstructured data, like images, voice, video, and combining with all the legacy data that the companies have been holding in their archives, is very powerful.
Megan – Can you share an example of how technologies like BQML or other Predictive Analytics tooling native to the GCP platform has unlocked some very quick potential customers?
Piyush – I will share with you an example of a client where they wanted to move their HR data onto the cloud. The state and local government client had about 12 different data sources, which needed to be quickly integrated into a Data Warehouse on Google Cloud Platform. When we started the project, the first challenge was to understand all of the data sources as well as combine them. Using advanced Predictive Analytics and Machine Learning, we were able to quickly understand the risk of attrition, who are the employees likely to leave. In a legacy environment, this type of analysis would have taken many months to build the infrastructure and many more months to configure the reports. This is an example that should get people excited.
Megan – Well, that’s fantastic, Piyush. We can’t wait to get you back on to talk about some more of these exciting examples to inspire our listeners about what they can do once they get their data into the Cloud. Thank you so much, Piyush
Piyush – Thank you, Megan.