Cloud Hosting

Ram says he is excited by the applicability of disruptive modern data architectures such as the data lake house, data fabric, and data mesh. He says the pendulum will continue to swing between centralisation through data lakes and lake houses, and the impending decentralisation with data mesh.

But what does this mean in practice? We all want our organisations to be data-driven but in the midst of a flood of data, and lots of noise, it can be hard to know how to get started, and what the marketing terminology all means.

You can watch and listen to our conversation here.

Ram explains he is visiting Australia to focus on customers, of which Cloudera has many valuable customers across a wide range of industries including telco, financial, government, and more. Ram notes that as 5G adoption grows in the telco space so too will more data be increasingly generated, so telco is particularly a big customer for a business like Cloudera.

Ram specifically mentions Deakin University which has adopted Clouderas products to find new ways of interacting with students and providing a greater personalised experience for them.

Ram also spoke about the Australian Defence Force, which is rolling out a system around personalised medicines on top of Cloudera.

I put to Ram that people say, "data is the new oil, and questioned if this metaphor was appropriate. Ram humorously responded with a resounding yes! Data is the fuel that powers new ways of doing business. Theres a lot of value, he said. But like oil, bad things happen if there is a leak or breach or spill. You have to be careful how you manage and handle it, and then you can monetise it and make new customer experiences.

Data is the enabler," he said.

That brought us nicely to the issue of challenges. Companies want to extract actionable insights out of their data, but most organisations will not have their data nicely categorised or packaged.

"What causes a lot of complexity is that data is everywhere, Ram said. It used to be in nice transitional systems that you got a batch report out of once per day. Now data is in social media, the cloud, SaaS applications, catalogue feeds from outside companies - there is so much data and so many kinds of data that being able to handle all consistently is a big challenge.

However, there is hope. Firstly, Ram explains, you need to think about data management. Regardless of whether data is on-premises or in the cloud, you must think about your data management and treat it as the same from a security standpoint so you set your policies once.

This is where the data fabric comes in; "the whole point is to make data management consistent, regardless of where the data is under the fabric, Ram explained.

Here is where a product like Cloudera comes in; it effectively implements a fabric that can work with hybrid multi-cloud so it does not matter where your data is located. Cloudera will manage it consistently.

Having the right use of data is very topical, Ram says. You want it accessible, but only to the right people. Here, he says, Cloudera can make a big difference - and just like cybersecurity, protecting data access must be a defence in depth. One set of controls around who can access it, another on encryption, another on personally identifying information, others on masking sensitive data further - you must use a variety of techniques.

This is like "having both a belt and suspenders so youre never caught without your pants, Ram said.

Ram also says the way we deal with complexity is to put data in the hands of the people who need it.

"Five to six years ago IT teams would interact with a product like Cloudera, he said. Now as more companies are data-driven, they need thousands of people who can store and manage data.

This is why it is critical to meet people where they are, he explained, and to provide a big range of tools. SQL is important, but now we have notebooks, machine learning, and self-service data analytics. If you have a consistent experience thats good, but you need to put it in the hands of people to make decisions. You must enable practitioners to get access to the data they need using the tools they know.

When it comes to 2023 and beyond, Ram says the data mesh is something we will increasingly hear about. It comes down to time-to-value. If you could do something in three months last year, you now need to do it in three weeks this year. You have a need for speed, to quickly take data use cases from concept to production, that all directly leads to business outcomes.

Ram also noted that data sovereignty in Australia and the rest of the APAC region is ahead of where the USA is today. As individuals, we all care about how data will be used, and this is a trend where complying with regulations helps us build systems more responsive to what the customer wants.

Finally, Ram spoke about ChatGPT and how amazing it is that this has so quickly become part of our conversation. Ram says this way of interaction is very new and impressive, with a language model following along with your conversation. Conversational analytics will become very compelling, he said.

"For us in the data world, our tagline used to be 'ask bigger questions - ChatGPT is allowing people to ask whatever they want in a conversational way and they will want the same from their enterprise data, Ram said.

"Once an interaction pattern becomes exposed we learn from it, but put in the guard rails so that applications like education and healthcare have precision as well as conversation.

"That's what's nice and exciting for me about the data space. We wouldnt have been having this conversation six months ago, Ram said.

See the article here:
Cloudera CTO Ram Venkatesh speaks with iTWireTV about the ... - iTWire