Autonomic architectures should top CTOs’ priority list: Mindtree report

Bengaluru- and New Jersey-headquartered technology solutions provider Mindtree has released the fifth edition of Tech Beacon, an annual report that provides technology guidance for enterprises.  

Tech Beacon suggests technologies for enterprises to invest in, experiment with or watch closely for future applications, what the company also refers to as the ‘Invest, Experiment and Watch’ strategy.

Mindtree said that the three classifications were based on the maturity and production readiness of the respective technologies.   

“Today, with aggressive digital disruption proliferating, the global business environment is transforming faster than ever,” said Madhusudhan KM, Chief Technology Officer, Mindtree. 

Madhusudhan also said that Tech Beacon offers guidance to enterprises based on industry requirements, changing business dynamics and the anticipated road of technology evolution.

Some of the key recommendations that the report stated pertained to advancements in systems interacting with systems, turning information into insights and bettering interactions with humans.

In terms of taking forth machine-to-machine interactions, more specifically the scalability problems with micro-services, MindTree said that it could be addressed with distributed architectures and service mesh owing to the consistent discovery, security, tracing and failure handling service capabilities of the two technologies.

The company also stressed that reactive autonomic architectures should be at the top of the CTO’s priority list, as it simplifies the end user experience though smart predictions in uncertain business environments.

In terms of getting better insights, the company said that machine learning was limited to the processes involving data harvesting while humans are still used for manual tasks such as data cleanup and outlier removal. Mindtree said that the need for integrating exploratory data analysis as a part of the processing pipeline could be accomplished by utilising Lambda and Kappa data processing techniques to generate useful insights from the available data.

“Lambda offers a consistent pattern and a real-time view into the generated insight, storing the data and then generating the insight in the batch layer,” the report said.  

While Kappa would work as the net iteration in this evolution to generate insights as and when the data arrives into the streaming layer.

In the area of improving interactions with humans, organisations still increasingly use conversational systems as a channel to enable machines to interact with humans. The report stated that the enterprises should use libraries such as Rasa, Wit.ai, and Microsoft's LUIS to enhance their core language processing systems and enable more humanised conversations with existing systems.