ITechnology Interview with Scott Toborg, Director of Product Management Data Science and Advanced Analytics at Teradata

“AI/ML will see greater scaling as enterprises finally figure out how to put the technology into full-scale production across many different application domains.”

Hi, please tell us about your role at Teradata and how you arrived here?

Hi, my name is Scott Toborg. I lead product management for our data science and advanced analytics products. Prior to this role, I was a Teradata Field Data Scientist helping our Fortune 500 customers implement Teradata Analytics to solve critical business problems.

Prior to Teradata I worked at AT&T as a Data Scientist in their Marketing Analytics group using Teradata Analytics to build predictive models for customer churn, cross-sell, and acquisition. I’ve worked in data science for nearly 30 years starting with research in machine learning for computer vision/image processing at HRL Labs, AT&T Labs and the MCC Research Consortium.

It’s been two years since the onset of the COVID pandemic. As a technology leader, how would rate the performance of the countries in fighting the pandemic with positive outcomes? Could we have all done better with suitable utilization of technologies and resources?

Absolutely. During the pandemic, we saw newer applications of even existing technologies.

For example, during COVID, video conferencing was not only primarily leveraged for work meetings, but it was heavily leveraged for virtual doctor visits, as well. Patients with symptoms did not need to visit the doctor physically, thereby decreasing the chances of spreading the virus.

Additionally, AI/ML-powered chat bots were further enhanced to carry out initial assessment of patients based on local and national health guidelines. This reduced the waiting time for COVID patients all while providing a more personal experience with natural language processing (NLP) baked into the bots.

In the last two years, even though technology found newer ways to help reduce the impact of the pandemic as evident above, wider adoption of such use cases by local, state and federal institutions could have reduced impact even further.

Read More: Mini, Micro and Nano Apps: What 2022 Looks like for IT DevOps Industry

Tell us more about the recent integration of Teradata Vantage data with Microsoft Azure. How do you bring together Big Data, Cloud Computing and AI Ops within one roof with this integration?

We are working with Microsoft Azure to leverage the best of both companies. Teradata can play a strong role in data ingestion, integration, and preparation (the most time consuming and costly part of analytics). While Azure ML has strong capabilities for advanced modeling. We are opening our platform to enable straightforward and easy integration of our native analytics together with open-source and cloud-native analytic services.

How would this unique umbrella technology integration improve data analytics workloads across an enterprise? What kind of customer experience and security reporting platform are you promising to your joint-customers?

We see many customers moving their AI/ML workloads to the cloud to take advantage of the flexibility that cloud computing brings by launching data discovery and model experimentation activities, and then scaling the final model to fit varying production needs. By having data management, cloud computing and AI/ML “under one roof,” customers can minimize data movement, decreasing costs and time to value. Teradata provides a low-risk migration path to our joint-customers to modernize data analytics on Teradata Vantage and Azure.

Read More: ITechnology Interview with Francis Carden, VP, Intelligent Automation and Robotics at Pega

What kind of challenges do organizations suffer due to their ever-growing data analytics workloads? How does Teradata solve these issues?

While moving to the cloud offers tremendous flexibility, it does not necessarily reduce costs. Customers must be vigilant in monitoring and managing cloud costs for consumption-hungry workloads like advanced analytics. Teradata helps by optimizing workloads to minimize costs. In fact, customer benchmarks consistently show Teradata with the lowest cost per query of any other system for large workloads. We advise customers to leverage Teradata’s built-in analytics for data preparation and feature engineering, and then, if desired, use open-source or cloud-native modeling services. We can then import the models back into Teradata Vantage for parallel execution in production.

Do you support evaluating cloud strategies related to modernization to solve data analytics challenges?

Yes. Many companies confuse cost savings as the primary reason they are moving to cloud. However, the primary promise of cloud is the agility it brings to businesses to move at a much rapid pace vs. on-prem. They can prototype a new idea very quickly in cloud, fail fast if the idea is not working out, and move on to the next idea. Modernizing an analytics ecosystem is key, and in most cases, mandatory before the agility benefits of cloud can be realized.

A critical challenge is the adoption of AI across an enterprise, especially in Cloud-based data analytics and visualization. What are your predictions for the AI-based Cloud marketplaces in 2022?

Yes, we agree. We see cloud-based analytics as making this adoption easier. Customers can spin-up a Sandbox for experimentation to prove-out an AI project. If it does not work, then the instance can be shutdown with minimal overall cost.

We predict that cloud adoption will continue to grow at a significant pace fueled by the needs of AI/ML. Cloud migrations and expansions will be justified and initiated for the purpose of implementing AI/ML applications. AI/ML will see greater scaling as enterprises finally figure out how to put the technology into full-scale production across many different application domains.

Which industries and business verticals within the organization are benefiting the most from AI’s adoption in Cloud and data analytics?

The industries that naturally generate enormous quantities of data were the first to deploy and reap the benefits of AI/ML. This includes retail, telecom, finance, healthcare and life sciences. We predict IoT technologies will spawn many new use cases for AI/ML in the areas of manufacturing, energy, transportation, and automotive.

Read More: ITechnology Interview with Cindy Provin, SVP and General Manager, Identity and Data Security at Entrust

Tell us more about your vision for advanced AI applications, especially in the emerging verticals within telecom, healthcare and life sciences.

We already play a significant role in all of these industries. Our vision is to provide a large-scale analytics production platform that will enable customers with the suite of tools and capabilities to drive value for their most important business problems. In telecom, this includes applications like real-time streaming to support 5G network diagnostics, customer alerts, or service optimization. In healthcare we imagine supporting customer/doctor recommendations, analysis and prediction of costs, or assisted diagnosis. As we have recently seen, life sciences has already benefited greatly from AI/ML’s ability to assist in the discovery of new d****. We also see AI/ML assisting in other critical areas like drug supply-chain optimization.

What advice would you give to every CIO/ CISO looking to invest in Cloud solutions?

I recommend thinking about cloud solutions with the following framework:

• Be clear in your motivations in moving to a cloud solution. “Everyone is moving to cloud” is not the right motivation. You must know what challenges your company is specifically facing on-prem that can be solved by a cloud solution.
• Plan your migration with a cloud solution vendor that can support your migration journey without adding risk. For example, because Teradata Vantage supports a hybrid architecture (on-prem and cloud), enterprise-scale customers can migrate at their own pace and continue to operate in a hybrid model per their business needs. We’ve seen large enterprises take multiple years to move all their applications and associated data to cloud, and Teradata closely partners with them in this journey.
• Security is key everywhere whether on-prem or cloud. With that said, cloud is a shared responsibility model. You will not get as much control over everything in cloud as you did on-prem. That is intentional in most cases. Understand and prepare for the shared responsibility model better as it is both a technical and a cultural shift.
• The end of a migration is not the end game. Once migrated, evaluate what components of the analytics architecture can be further modernized to realize more benefits of cloud. Prioritize the list in terms of strategic and business impact and continue modernization. Innovation in cloud never stops!

Thank you, Scott ! That was fun and we hope to see you back on itechnologyseries.com soon.