How to turn automation into success

Posted: 1 November 2017 | By Charlie Moloney

Automation frees up employees’ time for high-value tasks by liberating them from mindless chores, which should mean that enterprises can do more with the same resources.

Traditionally this has required single-purpose machinery to do mindless tasks with as much complexity and subtlety as the steam powered hammer that the mythical John Henry competed against.

But when automating a modern IT system, one-trick technology won’t fit the bill.

When dealing with such scale and complexity, technology must be able to think, and not only that but it should be able to think like an expert.

Why has automation not been delivering results?

“The whole approach of robotic automation is driven on the premise that what you automate is not changing”, said Dr. Harrick Vin, a Vice President and the Global Head of Digitate, a Tata Consultancy Services (TCS) Venture.

“Unfortunately, this assumption does not hold true in most enterprises”, Dr. Vin explained during an Access AI webinar.

“There has been an acceleration of the pace with which change is being introduced in an enterprise.”

“As a result, what was a standard operating procedure yesterday stops being relevant today. This essentially leads to a rapid automation obsolescence, making it difficult to sustain automation benefits over time.”

Thinking like an expert and taking the initiative

Business is accelerating, and the speed of decision making must accelerate along with it.

With an app or an API for everything there is abundant potential for automating anything from anywhere.

“Enterprises have to harness [this potential] to increase differentiation and to increase innovation”, Dr. Vin told us.

“Technology must be intelligent, in the sense that it has to have the ability to take decisions, and it should be autonomous enough to operationalise those decisions in real time.”

Dr. Vin’s team at Digitate have attempted to solve this problem by creating a new technology: Ignio.

“We have created an intelligent virtual expert”, Dr. Vin told us, “It thinks like an expert because we have taught Ignio domain knowledge about a wide range of common technologies that are present in an enterprise.”

“With Ignio, we combined the best of machine learning AI with knowledge representation about the domain and the ability to infer and construct procedures dynamically using exactly what a human brain is doing, which is combining context awareness with patterns and skills.”

I am Ignio, and I am a learner

The Ignio ‘studio’ has been built to capture knowledge about the domain in which a solution will be applied, and by data mining it can understand its environment and how it all relates together.

For example, Ignio understands the concept of storage and storage volumes; networks and network devices; hypervisors and virtual machines; and databases and article databases.

And it understands that in an article database of SQL queries, SQL queries are resident on tables, tables are stored in table spaces, and table spaces are stored in files, which are stored in volumes, which are stored in hard disks.

“In a sense Ignio has become a cross-functional expert which understands the facts, entities, and relationships about a large number of technologies,” said Dr. Vin.

“It has the ability to now tap into different data sources, and mine information about the context of an enterprise.”

Where is Ignio deployed, and what has it achieved?

A major retail bank is using Ignio to understand their entire batch environment that runs 100,000+ batch jobs across many different schedulers and platforms.

“The bank is using Ignio to predict the batch run before it begins”, Dr. Vin said, “and starting to do real-time adaptation of prediction so you can predict what is likely to fail, then take proactive action, thereby converting essentially the reactive to proactive, and thereby significantly improving the resilience of the business.”

“Most organisations take days to onboard an employee. With Ignio onboarding can be done within less than two hours end to end.”

In most cases customers have seen an up to 90% reduction in mean time to results in day to day activities.

A good example is customer onboarding where, due to the number of systems that get touched, most organisations take days to onboard an employee.

With Ignio onboarding can be done within less than two hours end to end.

“No one person within an enterprise has an intimate understanding of everything”, Dr. Vin told us, “Ignio basically taps into a whole bunch of data sources and schedulers, builds a global view, and then uses that to understand and predict.”

Ignio is no steam-powered hammer, but it demonstrates the same point that machines can outperform humans in routine tasks.

But what it also shows is that a machine that can think, understand, and advise humans will be more valuable in the long term than one that humans have to lead by the nose.


Related industries

Related functions

Related topics

Related organisations

Related key players