Bengaluru: Jim Goodnight, co-founder and CEO of the $3.1-billion analytics company SAS, said analytics models are getting much, much better because of the ability to enhance them with AI techniques. Image recognition and voice detects are among the major uses of AI, he said in an exclusive interaction with TOI.
Google and Apple, he noted, are doing a lot of the voice detects. “I use Google all the time to ask questions. And they spell every word I say correctly. So I am amazed how good their technology is going. We are doing a lot of work in image, trying to detect whether or not tumors, liver tumors are getting smaller because of chemotherapy. We have seen a lot of the uses of medical imaging to recognise whether a disease is present or not. We are using imaging also on the assembly line to determine whether a part is defective or not,” he said.
For Lockheed Martin, SAS systems take all the sensor readings off their C-130 planes and are able to forecast when parts are going to malfunction so that they can go ahead and repair them before they malfunction. “We do the same thing for Volvo trucks. Volvo has a computer on board every truck and we take the readings from that and then forecast when maintenance is going to be needed. We do anti-money laundering for over 200 banks right now around the world,” Goodnight said.
Right now, he said, the company has got a lot of work on the Covid situation, trying to help a number of governors and cities help monitor it and forecast how much longer it would go on so that they can figure out the need for intensive care beds, ventilators and personal protective equipment. “We are also doing a lot of work with contact tracing, where we follow people who have been around a person who has Covid,” he said.
Decades before data analytics became the buzzword it is today, SAS was doing analytics. Goodnight, now 77, co-founded the company in 1976 with three other faculty members of North Carolina State University, and he has been the CEO since. But SAS solutions were then affordable only for bigger companies. And it didn’t have the AI/ML techniques now available.
Today, a major effort is to, as Goodnight calls it, democratise analytics – “make analytics easy enough to use and easy enough to understand so that just about anybody can do it.”
So, how easy is it today? “Well, it is just one click really,” he says. “If there are variables that you want to include in your models, you can ask the system to automatically try hundreds of different models. That automation process is extremely useful. Because it allows people that are not modeling experts to try all sorts of different models. And we will recommend one that we like the best.”
Goodnight said the company does a lot of R&D in its Pune centre. “We have some very, very smart people in Pune. They manage entire projects,” he said.
SAS recently partnered with the National Health Authority (NHA), the government body implementing the insurance scheme Ayushman Bharat. SAS will enable data analytics to help monitor fraud and abuse arising anywhere in the implementation structure of the scheme.