(US and Canada) Kiran Kanetkar, now Pendulum Therapeutics VP of Data & Analytics and former Petco Senior Director of Data Engineering, BI and Analytics, speaks with David Mariani, AtScale CEO, about self-service tools, the data warehouse versus data lake dilemma, and the need to implement data ops techniques.

Kanetkar starts by talking about the tools being considered at Petco to deliver self-service analytics. He says that the organization is looking at open source technologies. It is currently using Python primarily for data analytics along with the tools that go along with it. He mentions that the team keeps track of open source tools and tries to provide their capabilities to business users.

Next, Kanetkar goes on to discuss the impact of on-prem infrastructure versus the cloud on the delivery of self-service analytics. He explains that on-premise infrastructure will always have scaling problems with the influx of vast amounts of data. On the other hand, he says cloud storage is cost-effective and the computing power can be scaled up or down depending on the requirements. Cloud is the way to go when it comes to analytics and the scale for computing power, he suggests.

Sharing his take on the choice between a data warehouse versus a data lake, Kanetkar says that data warehouses come with such a scenario that the concept of using a data lake is taking precedence to enable self-service.

Kanetkar further stresses that the adoption of cloud data ops is a key requirement moving forward. He elaborates that the cloud works on a pay-as-you-go model, not having automation to shut down the servers once users are done with processing, which can attract hefty costs. Automation capabilities can only be built using data ops methodologies, Kanetkar concludes.