Automation to bring-in significant efficiencies in data center operations

Data centers (DC) have become akin to essential services in the times of the pandemic, when WFH gained momentum and continue to remain as a mode of working in the industry. Thus digitisation gains momentum and cloud is emerging as an attractive proposition for enterprises of all sizes. Bajaj Allianz General Insurance has reaped benefits of moving to a cloud environment. The company is currently under the process of moving to a public cloud environment. 

The Bajaj Allianz General Insurance has not only transformed the core insurance platform but is also one of the large general insurance companies to have moved the core insurance system onto the public cloud. The company is currently working on a hybrid model and moving towards a public cloud model. The move will transform the way data resides in the information systems – from siloed data marts to a comprehensive data lake. But what does it mean for the business? There is complete visibility about the customers, on which products they own from the company for e.g. the customers owning a home, motor, or health insurance from the same company has to be made visible on a single screen, which not only has a direct bearing on the success of the promotions and pre-approved products sent out to them, but also the ease with which the customer can be serviced. “This has helped the company immensely in the last few months,” says KV Dipu, Head – Operations & Customer Service, Bajaj Allianz General Insurance.

DC automation is a big area to explore given the reliance on manual intervention required to man the DC operations. DC automation is being considered by many companies now given the steep learning curve in getting up to the speed with the Data Center Infrastructure Management (DCIM) solutions, which are in many ways complex to manage and require expert overview. “This was all managed on spreadsheets and management protocols, which can now be replaced by IoT (sensors) and AI. These tools can complete DCIM tasks in a far more real time and efficient manner,” says Dipu. AI is a natural corollary to manage various DC operations because of the data proliferation from public cloud adoption. We are in the world of Big Data now. As a result the hardware constitution has changed – be it the chip size, processing units, algorithms, machine output, the instrumentation, etc. On the other side, AI will also bring in efficiencies on some of the other components of the backend value chain, be it the building management system, branch circuit monitoring systems, sensors, environment monitoring system, power management system or DCIM.

The IRDAI recently allowed insurance companies to use the regulatory sandbox for experimenting with specific insurance products and solutions. The pre-condition to that is the preparedness of the DC environment coupled with data, application and the procedures in place.

The DC model adopted ensures a smooth sandbox experimentation. Giving a 360 degree view of the customer data made by dismantling the siloed data marts into a data lake prepares the ground for making multiple regulatory sandbox testing possible to be conducted at a rapid pace and thus giving a first mover advantage to the company.

An efficient data management strategy in any enterprise makes it ready to accelerate the product testing in a sandbox environment and then go for launch, which is ideally the purpose of the regulatory sandbox is i.e. to shorten the product launch lifecycle by not going through the rigmarole of the regulatory hurdles but testing on a closed user group environment followed by a full blown launch while also adhering all regulatory requirements. “The point here is to accelerate product launches but without compromising on the necessary controls in place,” says Dipu.

There are a plethora of options available for DC adoption, from the data centre developers and cloud hosting providers. The following points can prove to be prerequisites for ensuring not only making the right choice of adopting a particular model of DC adoption.

– An apt strategy around DC recovery and DC access
– Company’s suitability for adopting a particular DC model
– Strategic decision on inducing AI in DC
– Correcting the poor documentation of the critical processes
– Preparing a step-by-step, schedule and plan of the configuration procedures because there is a possibility that some systems may be obsolete
– Calendarise activities in coordination with various teams to make sure there is minimal impact on applications and various controls on security and access
– Plan for data in mission critical situations, like the pandemic.

Going forward, general insurance companies should prepare themselves for data localisation regulation; investing more in DC security and empowering the customer to have a complete view of the relationship with the company and thus seek or explore options on how GI companies can better service them.