Photo: Abhas Ricky, Chief Strategy Officer at Cloudera. Credit: Courtesy of the interviewee.
By Patricia Ruiz Guevera
Hybrid cloud data management and analytics is a growing sector whose offering and adoption has increased at a remarkable rate in recent years. In terms of infrastructure, we have the well-known storage usage but, if we keep looking amongst the clouds, we can see a whole lot more: computing capabilities and software services. Above all it is data that is the centre of this particular universe.
If we’re talking about companies that offer hybrid cloud data management, then Cloudera intrinsically carries the concept of cloud computing in its name. Since 2008 this American company has been exploring different solutions to handle huge volumes of data that are constantly growing in real time, in a way that could be easily transferable to all businesses and industries.
To really get a grip on it through real life cases - from the health sector through to insurance companies, and also taking in telecommunications - we spoke with Abhas Ricky, Chief Strategy Officer of Cloudera, who reminds us of the importance of taking a deep dive into concepts and not just floating around the surface: because cloud computing covers a huge area.
Why is cloud is such a big advantage for companies these days?
Cloud computing is hugely powerful for three important reasons.
The first is time. A couple of decades ago a business would have to get a server, some months later the software to go with it and then, after even more time had passed, start to try it all out. Nowadays in cloud computing, everything is a click away: if I’m an application developer, if I’m an end user or if I’m a business person, I can try a product and quickly work out how useful it is.
The second are services. This improvement encompasses flexibility, mobility and an increase in collaboration, but, above all, continuity thanks to its seamless integration. Quality control, loss prevention, disaster recover procedures, automatic updates… All of this is combined - which means that a business can be scaled much faster.
The third is cost. The main benefit of cloud computing is that you pay for what you use. The majority of the pricing models are structured in a way that a user consumes a certain number of credits and you don’t need to get lots of different programmes for each type of use.
These three advantages seem like they could be of use to any business, regardless of sector.
Yes, if we look at cloud computing use, we can see that no other technology in the past 30 years has grown as rapidly. Its annual growth rate is more than 20-22%. This is thanks to the fact that it has been adopted at face value by lots of industries. Whatever the sector, although usage may be different, the benefits are similar.
Despite such glowing reviews, has there been some reticence in its adoption?
Before its use became so generalised, there was a bit of concern in some sectors that they might get trapped - that they would be tied into a service in order to appreciate its benefits. But people soon realised that the advantages far outweigh the downsides. It’s also important to understand and differentiate about which cloud we’re talking about - since cloud computing is a huge term. It’s crucial to understand and separate the nuances instead of using ‘cloud’ as an umbrella term.
Public cloud, private cloud, hybrid cloud - what’s the best?
The majority of clients in the majority of industries are changing their focus to a hybrid cloud system, both from a costs perspective and from a security one too. In Europe there currently exists a concept called ‘Cloud Computing Concentration Risk.’ For example, the majority of banks have their treasury management systems and their accounting books in three main cloud providers - and this carries with it a huge potential risk. These strategies started out in a public cloud, so nowadays they are diversifying to hybrid cloud systems.
To really understand the benefits of cloud computing, it’s best to look at successful case studies in different industries. What project would you highlight?
The Spanish company of TIC services management, Zelenza, company was facing an increase in the volume of its data and did not have the technology capable of managing, analyzing, and scaling it efficiently for decision making. The team considered the more traditional option to use relational databases but were concerned that it could not handle the anticipated data volumes. The company turned to the Cloudera Data Platform (CDP) to manage data intelligently, in real-time, to make behavioral predictions or identify possible saturation points in the network, helping to continuously improve the quality of the network and the customer experience.
And how can this be applied in the telecoms sector?
Constructing an intelligent mobile network with an advanced management of data requires offering knowledge that frees up processes and really underpins effective decision making.
We’ve worked with a large telecommunications provider which had a huge short-coming: they didn’t have a centralised repository to efficiently manage, analyse, identify and scale decision-making processes in real time; therefore making behavioural predictions or identifying possible saturation points in the network wasn’t easy.
So their challenge was to construct infrastructure which could sustain all this critical information. They chose the CDP public cloud and secured the extraction of data from 144,000 cells of mobile technology. This massively improved their service costs since they no longer had to analyse data semi-manually plus it improved their efficiency, quality and forward-planning thanks to the proactive analysis of behaviour that they can now carry out.
And a third example from the insurance sector?
One of our clients is a really large European insurance multinational that is building a centralised learning system - almost like a type of Artificial Intelligence factory. During the development stage, they are allowing teams of data scientists to share use cases through the cloud. Therefore they can create decision-making processes almost in real time.
Their main challenge has been the incorporation of the data scientists into the infrastructure of the cloud in the most efficient and quickest way possible. In terms of clients, we’re talking of between 10 to 12 million, there’s an immense quantity of data to manage. So they also chose our CDP and our Cloudera Machine Learning [CML] ecosystem, an automatic learning product in the public cloud. Thanks to both, the teams can work in an agile way and get connected in a question of minutes.
In these use cases there’s a mixture of Artificial Intelligence, Big Data, automatic learning and advanced computation. How has cloud computing achieved this synergy between different layers of technology?
The central piece is data and its quality. Data governance, metadata management, master data processing, data modelling, data cleaning, data profiling… Any one of these categories is a business in itself. The emerging architectures imbibe this data and it’s crucial that enough attention is paid to maintaining its quality.
“It’s crucial to increase productivity between the data teams, as data is a team sport”
For example, autonomous vehicles, health mapping, geospatial transport analysis... the majority of these businesses are labelling data at a halfway point between manually and AI. Is it a plastic? Is it rubber? Is it a support? We need to understand what the data represents as this will be directly related to the processes. It’s crucial to increase productivity between the data teams, as data is a team sport.
Since Cloudera is a technology and solutions provider, how would you say that the cloud can help clients in particular?
We have a discovery session with our clients and ask them some key questions. What problems do they have in their business? What are their main pain points? What financial objectives do they want to achieve? With all the answers at hand, we outline a set of use cases and start to bring together the sources of data so that we can work towards a solution. Some clients already have system integrators in place but others don’t - so we have to do a bit of hand-holding through the process.
How do you foresee the future of services in the cloud?
We’re only really scratching the service of what the cloud can offer. Right now the global market in public cloud infrastructure is expected to grow to US$120 billion. Meanwhile research and advisory company Forrester have predicted that this sector will grow by 20% annually.
The hybrid multi-cloud is a key term that is bound to dominate the future. Gartner recently released a report which highlighted the importance of providers adopting a hybrid model approach to avoid potentially losing work in the future. I believe the multi-cloud will be imperative in enabling processes to be moved between the private and public domains in a fluid way.
“We’re only really scratching the service of what the cloud can offer. The hybrid multi-cloud is a key term that is bound to dominate the future.”
There is also a lot of talk about new frontiers like the edge. Lots of people say that the edge is the new cloud in certain aspects and could take off thanks to 5G and the engineering of artificial intelligence - which is a key component. There will also be a great expansion of use cases and software development.
Lastly, computing without a server - the true paradigm of cloud computing - is here to stay.
Wherever you look, in whichever industry or region, we’re seeing all sorts of cloud usage. There hasn’t been another technology which has grown at this rate in the last 30 years and I don’t think anything can compare in terms of investment that it has received, either.
Published by OPINNO © 2022 MIT TECHNOLOGY REVIEW spanish edition.