In this interview from Oracle AI World 2026, Avinash Singh, vice president of technology services at Equifax, joins theCUBE's Dave Vellante to discuss how the credit bureau traded daily database firefighting for hands-off billing operations with Oracle Autonomous AI Database at AWS. Singh explains how Equifax's global billing system, which processes billions of transactions each month, outgrew the 65-terabyte ceiling of its previous database service. Since going live on Exadata-powered Autonomous AI Database in April, the team has seen SQL performance tickets drop from 15 to 20 per day to just one or two per month. He credits auto-indexing and an optimizer that eliminates manual query tuning.
The conversation also touches on the migration itself, which required zero application rewrites and was completed over a weekend with no downtime. With nearly 100 terabytes now on the platform, Singh details how statistics gathering shrank from three to four days to three to four hours. Usage rating improved by 200% to 300%, helping Equifax close its books faster. Configuration adjustments to harness Exadata also boosted batch processing from 100 to 200 rows per transaction to as many as 30,000. For the billing operations team, the change means fewer month-end fire drills, on-time invoices and fewer revenue losses from delayed billing. Looking ahead, Singh outlines how Equifax's cloud foundation and patented EFX.AI position the company to adopt machine learning and agentic workflows. He also notes that agentic AI introduces new machine-to-machine fraud risks that are reshaping the company's product strategy. From a nearly complete cloud transformation to AI-ready billing data, Singh offers a practical playbook for enterprises looking to modernize mission-critical databases without disrupting the business.
Almost there!
We just sent you a verification email. Please verify your account to gain access to
Oracle AI World 2026. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Register for Oracle AI World 2026
Please fill out the information below. You will receive an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for Oracle AI World 2026.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Avinash Singh, Equifax
In this interview from Oracle AI World 2026, Avinash Singh, vice president of technology services at Equifax, joins theCUBE's Dave Vellante to discuss how the credit bureau traded daily database firefighting for hands-off billing operations with Oracle Autonomous AI Database at AWS. Singh explains how Equifax's global billing system, which processes billions of transactions each month, outgrew the 65-terabyte ceiling of its previous database service. Since going live on Exadata-powered Autonomous AI Database in April, the team has seen SQL performance tickets drop from 15 to 20 per day to just one or two per month. He credits auto-indexing and an optimizer that eliminates manual query tuning.
The conversation also touches on the migration itself, which required zero application rewrites and was completed over a weekend with no downtime. With nearly 100 terabytes now on the platform, Singh details how statistics gathering shrank from three to four days to three to four hours. Usage rating improved by 200% to 300%, helping Equifax close its books faster. Configuration adjustments to harness Exadata also boosted batch processing from 100 to 200 rows per transaction to as many as 30,000. For the billing operations team, the change means fewer month-end fire drills, on-time invoices and fewer revenue losses from delayed billing. Looking ahead, Singh outlines how Equifax's cloud foundation and patented EFX.AI position the company to adopt machine learning and agentic workflows. He also notes that agentic AI introduces new machine-to-machine fraud risks that are reshaping the company's product strategy. From a nearly complete cloud transformation to AI-ready billing data, Singh offers a practical playbook for enterprises looking to modernize mission-critical databases without disrupting the business.
play_circle_outlineMulti-cloud infrastructure management across AWS, GCP, OCI, Oracle Fusion
replyShare Clip
play_circle_outlineResolved AWS RDS 65TB limit by moving to Exadata-backed Autonomous Database
replyShare Clip
play_circle_outlineEquifax zero-downtime billing migration to Oracle Autonomous AI Database on AWS using Exadata configuration and GoldenGate partnerships
In this interview from Oracle AI World 2026, Avinash Singh, vice president of technology services at Equifax, joins theCUBE's Dave Vellante to discuss how the credit bureau traded daily database firefighting for hands-off billing operations with Oracle Autonomous AI Database at AWS. Singh explains how Equifax's global billing system, which processes billions of transactions each month, outgrew the 65-terabyte ceiling of its previous database service. Since going live on Exadata-powered Autonomous AI Database in April, the team has seen SQL performance tickets...Read more
exploreKeep Exploring
What business drivers are currently shaping your technology decisions and strategies?add
What motivated the decision to migrate the legacy billing system (running on AWS RDS Oracle) to Oracle Autonomous AI Database on Exadata?add
How was the migration from the AWS-hosted database to Oracle Autonomous Database on OCI carried out, and how disruptive was it in terms of downtime, required code changes, and business impact?add
>> Welcome back, everyone, to Oracle AI World 2026 here at the Venetian. I'm Dave Vellante. Avinash Singh is here. He's the Vice President of Technology Services at Equifax. And the title of this segment is Equifax Simplifies Financial Operations with Oracle Autonomous AI Database at AWS. How's that for some vendor marketing? All right. Good to see you, man. Good to see you. Yeah. What a show, huh? Oh, yeah. This place is packed. Unbelievable. Big keynotes this morning. Oracle went right to the business, the heart of business Serious company. Before we get into it, Avinash, tell us about your role, Vice President of Technology Services. What does that entail?
Avinash Singh
>> So, Dave, as a part of the technology services, it expands across, so I look over the cloud infrastructure for enterprise data and financial services. So we are a multi-cloud company, so I operate for AWS, GCP, Oracle OCI, Oracle Fusion, so my portfolio takes care of all those. Cloud infrastructure.
Dave Vellante
>> Okay, so I'm an Equifax customer, everybody kind of is, right? Just about to get a new car, so I got to unlock my credit. But you guys, of course, do a lot more than that, and you have rich data. Definitely. What are the key drivers of your business right now that are affecting your technology decisions and strategies?
Avinash Singh
>> So, being one of the three largest credit bureaus in US, we are the point We are pioneers on migrating that whole our on-prem data centers to cloud. We almost achieve 100% mark. And having said that, as part of our multi-year strategy, we are expanding our investment towards patented EFX.AI, which is going to use for deploying new products, capabilities for Equifax business altogether.
Dave Vellante
>> Yeah, the industry has changed quite a bit over the years. Obviously, mobile, the internet, has affected that. And then that obviously drives, your data and AI strategies. I understand you had a billing and revenue management project. You had a database associated with that. What were the objectives that you were trying to achieve on that project?
Avinash Singh
>> So, Dave, before that, going to the objective, let me talk a little bit about this billing product. So this is Oracle billing product, very rich in processing, high processing usage, rating and billing and invoicing. So behind the scenes, the database is very, very highly transactional database, like billions of transactions we process per month. So the demand of that, that it should be processing fast. Our usage should be rated fast, accurate, and so that we can build our customers proper invoices on time. So we were using earlier AWS RDS Oracle services, the awesome service, but the way that continuous growth, what we have in Equifax, migrating all kind of billing system, legacy billing system towards that our global billing system, which is running on Oracle BRM, our growth was expanding and we hit at the bar of that 65 terabyte limitation. And meanwhile, that in AWS, when it was announced back in 2025 that Oracle is doing this multi-cloud strategy and then now Autonomous AI Database is gonna be part of that. So, which has petabytes of storage capability. Running on that engineered system of Exadata. So we thought of it's a good opportunity that to meet our growth plans and all that on the billing system, we can take this service. Well,
Dave Vellante
>> it's kind of ironic, because Oracle was sort of late to the public cloud. It took a couple of iterations. And now they have by far the most comprehensive multi-cloud strategy in the business. It's kind of crazy when you think about it. I guess sometimes it's not best to be first. Let's figure it out, and then, because everybody, every customer we talk to, who has multiple clouds. I want to ask you about Oracle Autonomous AI Database. What was it about that? Was it the autonomy, the autonomous part of it? Was it about the AI? we did a study that showed that migrating to the Autonomous AI Database was the fastest way to get on the AI curve and drive AI ramp. But then there's the autonomous piece. So what drove you to that decision? Was it the autonomy or the AI or both?
Avinash Singh
>> Oh, Dave, I think that the way that you explain, I think that you covered half of that, what I'm going to go to speak. So basically that both. So that's the beauty of using AI database of Autonomous Database dedicated what we are using, where we migrated, we gone live in April, and we are gaining a lot of power, autonomy of that Autonomous Database. So let me give you one example. So we were earlier having like 15 to 20 tickets per day to solve the SQL performance issues. We have to run the performance optimizer, runs for 30 minutes, 1 hour, and then we just pin our SQL profiles for, oh, this query is working bad. Look, let me put this SQL profile. After 2 hours, oh, the stat is gonna go to stale. Oh, the SQL profile is changed. So what happened that once we move in Autonomous Database, that real power of Autonomous Database is like auto-indexing.
Dave Vellante
>> Yeah.
Avinash Singh
>> Their performance optimizer is so advanced that we don't need to tune a single query. And so my ticket process, because it's a highly OLTP customized work and all that we do based on that our business. It is like 1 to 2 tickets per month. So just think about that. 15 to 20 tickets per day turns to 1 to 2 tickets per month.
Dave Vellante
>> Wow.
Avinash Singh
>> So that's awesome.
Dave Vellante
>> Yeah, yeah, it's amazing. What about, anytime I hear the word migration, that's like an evil word in your world. There's always complexities. the vendor selling the technology makes it sound so easy. But then there's all these other dependencies that you as the customer have to deal with. So how long did the migration take? How difficult was it? Did you have to freeze code? Did you have to rewrite the application? Because if you have to freeze code, that affects the business. Take us through the migration. How painful was it?
Avinash Singh
>> Okay, good. So, Autonomous Database is not like something that is just reinvented. it was always available with OCI. So Autonomous Database was always available as an OCI service. The thing is that the beauty of this multi-cloud service with AWS, that all that our application stack was running on AWS, I just need to switch from one type of database service to another. So there is zero code writing because it is just about migrating the data from one database to another database. So we use that kind of some Oracle-provided services like zero downtime, Oracle GoldenGate and all that all clubbed together. And we achieved actually this migration without any downtime. So we did that, chose that over weekends and we migrated. We have that GoldenGate setup and zero downtime along with that. And that migration strategy works so well. Zero application rewritten. And we using this technology, migration was smooth. So AWS and Oracle supported both on this migration.
Dave Vellante
>> So the business side didn't really see any interruption.
Avinash Singh
>> Oh yeah, so it is very transparent for business. Yeah, zero downtime for them.
Dave Vellante
>> How about the results? What have you seen?
Avinash Singh
>> Oh yeah, I just want one of the results I talk in our one of the questions, right? So 15 to 20 tickets to 1 to 2 tickets per day, 15 to 20 tickets per day to 1 to 2 tickets per month. Big achievement. And because we are more than 75 terabytes when we migrated, now we have almost 100 terabytes data sitting on this database. So just think about that because of highly OLTP, that stat gets stale very soon. So earlier with the previous service, it was taking 3 to 4 days for gathering the stats for high performance in that subsequent processes. Now with Autonomous Database, Trust me. Guess how many hours it should be taking?
Dave Vellante
>> So it's, I don't know, 24 hours?
Avinash Singh
>> No, 3 to 4 hours.
Dave Vellante
>> Really?
Avinash Singh
>> Really. So we— that Autonomous DB provides that one service called TPURGENT. We can set the 64 parallels, and then because of Exadata power, it is able to deal with these 64 parallels very efficiently. So my gather stat reduced from 3 to 4 days to 3 to 4 hours. So it is now not like that we are setting a particular time. To run the gather stats, it's every weekend we set. So my stats never go stale. So that is one of the biggest benefits.
Dave Vellante
>> And you're running a lot of batch jobs too, right?
Avinash Singh
>> Oh yeah, because billing system is batch job.
Dave Vellante
>> So how did it affect batch performance?
Avinash Singh
>> So one of the batch jobs, what I told about billing system, is rating the usages, right? We get billions of usage files. So our rating improved almost 200 to 300%. So we are able to close our books way before what we have been closing.
Dave Vellante
>> So, were there any application changes that you were able to make with the extra database capacity? What did you do with that free capacity?
Avinash Singh
>> Yeah, so as I was saying, Dave, right? though we don't need to rewrite the code, but to harness the power of Exadata, we looked over certain configurations. For example, earlier with that earlier commodity database, we were having like 100 to 200 rows processing per transaction. Now with this batch, we are doing almost 15,000 to 30,000 rows per transaction. So you can think of that. So that is adding up that all our usage processing performance 100 to 200%. So just definitely we did some adjustment on the configuration to harness the power of Exadata.
Dave Vellante
>> Avinash, what does this mean for your customers? So obviously you've got internal customers and then you've got your customers' customers. What does it mean for them?
Avinash Singh
>> So for internal customer, our billing operations team, earlier they were just firefighting on the system problems. Hey, this particular usage did not come on time, so can we just do the adjustments? Can we do that, rerun those things because it did not meet that certain date? So they have been creating a lot of tickets to my application management service team. It comes to our DBA team, and then it was always firefighting, especially on the month ends. Now it's all smoothed up. Billing operation is more dealing with the real problems of the end customers. So, and it's a billing system, right? So the benefit of the billing systems is that getting the invoices on time. So there is no kind of problems of invoicing and then rather than, then again adjustment overall revenue kind of, we can fix those problems related to revenue losses if that any rating and billing invoicing get delayed. So this is although that's a backend system, but behind the scenes, indirectly, it is giving a huge impact on overall our growth strategies.
Dave Vellante
>> Great. What's the future look like for you? What do you want to accomplish 12 to 18 months from now? What do you want to be able to say, maybe that you can't say today, Avinash?
Avinash Singh
>> So, Dave, as you know, one of our strategies to moving AI database is harness the power of AI. Now that Equifax is investing a lot on our patented Equifax.ai and the same journey that we just put our billing data to AI database. So I just wanted to harness all the AI productivity and capabilities. So our foundation was Equifax Cloud, which we set up as part of that, our first initiative. So back in '24, '25, we done all kind of cloud transformation. Now with this cloud transformation, it gives a lot of ease for us to utilize AI, machine learning, agent workflows. And also, if you look at that, nowadays, it is very evident that agent AI can, do that machine-to-machine fraud. So we have to look over our product strategy, right? It is reshaping the products.
Dave Vellante
>> Yeah, wow.
Avinash Singh
>> So we have to think through, and that's why Equifax is way ahead in the game. They already put that multi-year strategy, first do the cloud transformation, which is that whole AI adoption, then investing on EFX.AI patented. So they are on the right way to work on those product reshaping. So yeah, pretty much.
Dave Vellante
>> Wow, it's created a whole new flywheel for you guys. Avinash, thank you so much for coming on theCUBE. We appreciate it. All right. And thank you. Keep it right there. You're watching theCUBE special coverage here at Oracle AI World 2026, live from the Venetian. I'm Dave Vellante.