Zak Murad, the Chief Technology and Information Officer of research, ratings and data firm Crisil, drives product development and spearheads initiatives in risk management solutions, artificial intelligence, machine learning, and cloud. He was earlier the global Head of the IT Infrastructure and Operations at S&P Global Ratings (owned by Crisil’s parent, S&P Global Inc.). In an exclusive interview, Murad spoke about how Crisil’s digital transformation initiatives like migrating to the cloud, leveraging AI and data analytics, and the changing role of the CTO and CIO, among other things. Edited excerpts:
What progress has Crisil made with its digital transformation initiatives?
The digital quotient of the entire company is getting elevated. There is digitization of all components -- People, Process, and Technology. Agile transformation is a big part of what we are doing. Moreover, things that you used to do manually are getting fully automated. The type of products we are building is changing as well -- we’re moving everything to the cloud.
We’ve been at this now for two years and have another two years to go. Our cloud migrations will be done by next year. We are also launching a lot of SaaS (software-as-a-service) products and working on a platform approach to building things. Also, given the way you exchange data is changing, security is a big deal because every threat actor (hacker / cybercriminal) wants to get access and exploit. So, we’re spending an incredible amount of time ensuring that things are fully secure.
You are both CTO and CIO. How do you combine these roles from a strategy perspective?
Crisil is a multifaceted organization. We do a lot of things like ratings, benchmarking, and risk solutions, and also provide many services and push out a lot of products. So, in a way, we are very tech heavy company. I’m quite hands on as a CTO -- I go through the code with the team where needed, and then allow the team to structure and build. As a CIO, I look at the product; look at how the information is flowing; how to secure it at a macro level; how to align (the products and services) with business objectives; and how to leverage technology to push the kind of growth you’re seeing within Crisil. I also spend time with clients to understand how we should be building better products to maximize our client value. These two strategies (as CTO and CIO) are aligned essentially to optimally utilize resources such as time and money.
Speaking about technologies, how are Crisil’s investments in Machine Learning, Deep Learning and NLP (natural language processing), paying dividends?
We have a GAC (Global Analytical Centre) division that works on providing data to S&P Global, which involves a lot of extraction work. We have a Data Science team, and we use Machine Learning to help with the extraction of data from the unstructured documents that we receive, such as filings, etc. The other side involves Data Science. Since we have a huge amount of data from a ratings perspective, as well as our other businesses, we are looking how to derive more intelligence out of that data by using and building Machine Learning models. We are a data company in many ways, and machine learning has a lot of use cases for us within data processing. We also do a lot of risk analytics, and we build models to understand risk, early warning systems, and things like that. ML can be leveraged in these areas as well.
How do you view the maturity of your clients when it comes to adopting these newer and emerging technologies?
It’s varied. When I look at cloud adoption, for example, I don’t see a huge amount of maturity in India as yet. The move to the cloud has started to accelerate in India now. Maybe India is behind by a year or two, but it will catch up very quickly and we are very much focused on moving things over to the getting cloud and bring new offerings as SaaS.
When it comes to machine Learning, there are some companies that are doing some work, especially in Data Science. A lot of companies have built data Lakes (data repositories of raw data) but many are struggling with them, resulting in data swamps.
How should companies look at return on investment (ROI) from these emerging technologies?
Companies are not monoliths. If you take a large bank, for instance, it has many divisions and not every division and not every product or service they offer is going to be under strong regulatory control, or security control from a regulatory standpoint. When they ask: ‘Does it make sense to move to the cloud?’ (they should know if they are) moving to the cloud because they want to save costs, or want to make use of services, or because they want agility.
We are moving to the cloud to gain agility because we feel we can offer a lot more value to our clients by leveraging the capabilities that the cloud has to offer. So, it’s targeted towards revenue and growth, more so than anything else and the way to do that is to have product innovation happening much faster than you would be able to do when residing inside your own data centre.
That said, you can absolutely use a cloud service inside your own data centre and make it work, depending on the kind of application you have. If you are really smart and have the right kind of applications, you may be able to spend about the same on an external cloud, as you were spending inside your own data centre.
What improvements have you seen within Crisil?
Within S&P, I moved about 175 applications to AWS (Amazon Web Services) in about 8 months. We have been running there now for over two years and quite successfully. So, you have to understand your cost and fundamentally understand what is that that you are trying to accomplish, and then work backwards from there to say, “Okay, this strategy allows me to get back to my business goals.”
If you want to build a product and you have a platform like our ‘FulKrum’, you can use it as an accelerator to build new products. Earlier, if you had to build a low complexity or medium complexity product, you could build it from the beginning to end -- getting the servers, etc., in the traditional mode. But this could take 5-6 months to get it all configured; and the entire cycle could take a year to get a product out. With new platforms, I am able to do it in less than 3 months. So, you cut your product cycle down from 1 year to 3 months, and then offer it as a SaaS service.
This can also help us bring in adjacencies with other products. Say you have a product, and the client finds value in another product as well, it can allow for seamless integration across those for a good user experience. These are the things that create better value for the client. Many companies have many products, but they don’t integrate with each other. A platform approach helps in accomplishing these things.
Given that we are already talking about a potential recession in many economies, are you seeing companies reducing their IT spends?
I haven’t seen anything alarming as yet. There are definitely indications (of a slowdown) that I am reading in the news, but I haven’t had conversations on the service side with clients that has given me an indication – one way or another.
I feel as long as we continue to provide client value, we are going to be fine. And client value can come in many different ways. There is opportunity as people are looking to move things to India as products and ODC (offshore development center) process. We have capabilities around all these areas.