Artificial intelligence: the science fiction buzzword of the 20th century, and the tech buzzword of the 21st. While its ubiquitous appearance in marketing copy may have diluted its meaning somewhat, true artificial intelligence is already powering the products, services and customer experiences of some of the tech industry’s biggest players. 

At its heart, artificial intelligence is perfect for several clearly defined use cases. Among them: UX/UI personalization, automated data management and analysis of that data at scale.

These applications make AI particularly important to e-commerce and big data companies, along with insurance and fintech companies that can build predictive models on technology that analyzes huge datasets. These engineering leaders are at the forefront of AI’s evolution in their respective fields, and they spoke to Built In about the trends they’re seeing across the industry. 

 

1. Connecting Every Facet of E-Commerce to the Consumer Experience

Dr. Olly Downs - Zulily
Zulily

Who: Dr. Olly Downs, VP of Martech, Data and Machine Learning at Zulily

The company: Zulily is an online e-commerce retailer that sources discounts for users on a wide spectrum of products, from clothing to furniture.

The trend: “Our long-term vision is that actionable intelligent services enabled by advanced data and machine learning methods connect every area of an e-commerce business to the customer experience. We think of Zulily as a fully connected system which optimally intermediates customers and our vendor ecosystem. We’re exploring leveraging data science to connect sales trends, while understanding product availability across the supply chain in order to predictively and efficiently order, market and deliver inventory to each customer.”

 

“By using demand forecasting, we can enable added agility in our supply chain, giving us the ability to better plan warehouse resources, which is critical in e-commerce.”

 

How they’re moving the needle: “We built our own machine learning platform that allows us to rapidly develop, test and deploy machine learning models and business solutions built on them. Thanks to our culture of experimentation and driving progress, our data scientists and engineers can deploy models at scale, and in a matter of hours, increase throughput and data consistency by leveraging and contributing to open source technologies like Feast. 

Our business model is unique — we don’t buy inventory until our customers do — which gives us agility but can make shipping complex. By using demand forecasting, we can enable added agility in our supply chain, giving us the ability to better plan warehouse resources, which is critical in e-commerce.

Lastly, we’re still in the early stages but I’m proud of our current experiments in using computer vision and natural language processing techniques to quickly understand products and accurately populate product attributes — which is important because we launch thousands of new products a day. The more completely and accurately we understand products, the more easily we can connect the right customers with them.”

 

2. Automated Market Analysis At Scale

Nadine Chakar - State Street
State Street

Who: Nadine Chakar, Head of Global Markets at State Street

The company: Financial services and bank holding company State Street has been around since 1792, making it the second-oldest bank in the United States. It now uses institutional experience and technology to empower investors. 

The trend: “We are increasingly seeing the benefits of embracing emerging technologies like robotics, low-code/no-code applications to improve workflow, support automation and predict outcomes. We use a mixture of internally developed frameworks along with external partners and vendors to allow us to customize our offering, accelerate time to market and access the right talent mix.”

 

“These sophisticated technological capabilities harness decades of human trading experience across multiple teams.”

 

How they’re moving the needle: “State Street Global Markets frequently uses cutting edge quantitative techniques and advanced technologies across its business portfolio, which includes foreign exchange and our research platform.

State Street’s foreign exchange team has been investing in its low latency electronic market making capabilities and client-facing algorithmic trading. These sophisticated technological capabilities harness decades of human trading experience across multiple teams to include our global markets’ dealing desks, thought leadership from our State Street Associates research group, technical specialists and experience in platforms like Global Link.

Through our State Street Associates research platform, we have developed a range of AI-driven tools to help our clients forecast returns and manage risk. These include: natural language processing of thousands of digital media sources to detect bias-adjusted sentiment surrounding companies and markets; a machine learning platform to forecast currency returns and compare these forecasts to the views of our human strategists; and high-frequency economic indicators that are tracking the changing consumption patterns of consumers through the COVID-19 pandemic.”

 

Find out who's hiring.
See jobs at top tech companies & startups
View All Jobs

 

3. Robotic Risk Modeling

 Vikas Vats, Chief Analytics Officer at Verisk
Verisk

Who: Vikas Vats, Chief Analytics Officer at Verisk

The company: Verisk uses data analytics to help businesses in the insurance, energy and finance sectors navigate risk. Headquartered in New Jersey, the company started as a non-profit advisory in 1971 and now works with customers in more than 30 countries. 

The trend: “AI is the cornerstone of digital decision making in the industry, and the primary business trend is to use it in risk modeling. From collecting and interpreting visual, audio and other unstructured data in conjunction with traditional datasets to building adaptable models, the industry is speeding up workflows, providing consumers and businesses with faster turnaround times both in the acquisition phase as well as in claims processing. Digital workflows, customized experiences, faster contract interpretation, active risk monitoring, better understanding of in-force books are all trends that are being driven by AI technologies — and are front and center for us. We feel AI will be ubiquitous in the insurance domain. Other areas we are interested in is the application of AI models in modeling catastrophic events like pandemics, hurricanes, earthquakes and cyber events.”

 

We have been at the forefront of engagement between the industry and regulators, and taken the steward role of making all players comfortable with the adoption of AI.”

 

What’s your role in this trend? “We have been at the forefront of engagement between the industry and regulators, and taken the steward role of making all players comfortable with the adoption of AI in ratings and underwriting. We have also employed AI technologies in facilitating consistent and faster application of insurance regulations through our collaboration with state insurance departments, who have been very open to adopting AI technologies to evaluate forms and contracts.

The application of AI technology in streamlining claims workflows is a trend that has gathered strong momentum. Another interesting area of application is in the use of visual data for loss estimation, for detecting conditions of assets and structural features. We have not only helped our carrier customers in cutting underwriting and claims adjustment time, but also enabled them to respond faster in the event of natural disasters.”

 

4. NLP As You’ve Never Seen It Before

Stephen Edginton, VP of Engineering at Epicor
Epicor

Who: Stephen Edginton, VP of Engineering at Epicor

The company: Austin-based Epicor builds industry specific software for manufacturing, retail, distribution and services businesses. These include enterprise resource planning, customer relationship management, supply chain management, human capital management and enterprise retail software platforms designed to boost organizational efficiency. 

The trend: “The hot technology we are watching at the moment is Open AI’s new GPT-3 Language Model. This really takes natural language processing to the next level, with its 175 billion parameters.

What does that mean though, and why is it interesting for business? Imagine being able to take a brief description from an email and generate the structure of an engineered product or quote and have a reply to that email automatically in your draft folder. Imagine being able to talk naturally to your computer and have it automate that process of what you are after. For example, it could respond to commands like, ‘Create a new order for ACME based on what they returned last week’ or ad hoc queries like, ‘Show me all containers arriving from China in the next two weeks.’”

 

“These models require computing power well beyond what’s available to mid-sized or even large businesses.”

 

How are you contributing? “Epicor is moving the needle with AI in a few ways, one of which is Epicor Virtual Agent. We continue to evolve EVA to leverage skills and trained language models, but we have to train these and write the skills. GPT-3 opens up new opportunities and means we may not have to write the skills. These models require computing power well beyond what’s available to mid-sized or even large businesses — it requires a super computer of 285,000 CPU cores and 10,000 GPU’s, not to mention the data scientists and data to train on.”

 

Delight Your CFOHow to Reduce AI Computing Costs

 

5. Hyper Personalization Comes to Fintech

Brandon Rembe, SVP of Product at Envestnet | Yodlee
Envestnet | Yodlee

Who: Brandon Rembe, SVP of Product at Envestnet | Yodlee

The company: Chicago-based Envestnet builds wealth management technology. In 2015 the company acquired data aggregation company Yodlee, combining the two as Envestnet | Yodlee.

The trend: “Hyper-personalized digital experiences have already taken over consumer technology. Consumers have grown accustomed to the personalized experiences they receive through Amazon, Netflix and others. Because of this, consumers are also yearning for similar interactions on all of their digital experiences, including financial matters. Using AI, machine learning and data to create hyper-personalized experiences is now extremely important in fintech and this capability allows financial service providers an opportunity to meet the high expectations of consumers.”

 

“This product suite allows financial institutions, fintechs and advisors to build, implement and scale the hyper-personalized digital experiences that consumers require.”

 

How are you contributing: “On August 5 Envestnet | Yodlee introduced its new Insights Solutions platform, which will be available in the third quarter of 2020. This product suite allows financial institutions, fintechs and advisors to build, implement and scale the hyper-personalized digital experiences that consumers require. Utilizing Insights Solutions, financial service providers will have an opportunity to leverage a combination of actionable insights, unique peer benchmarking capabilities and personalized views to develop experiences that can help their customers with all facets of their financial wellness. This is particularly noteworthy during these uncertain times. Combining Insights Solutions with the power of secure financial data enables financial service providers to advance the financial wellness of their customers.”

 

6. Automated Data Management Is Table Stakes

Dave Russell, Vice President of Enterprise Strategy at Veeam
Veeam

Who: Dave Russell, Vice President of Enterprise Strategy at Veeam

The company: Veeam builds data management software along with disaster recovery and backup platforms for virtual, physical and multi-cloud infrastructures. Headquartered in Switzerland, the company has offices throughout the U.S.

The trend: “We believe the use of AI data management will become not only prevalent, but crucial to modern business success. In this space, true innovation isn’t just about capturing more data — it’s about having that data always available, responding to it and separating signals from the noise in order to improve processes and drive better business decisions. With AI and machine learning, decisions can then be made based on the patterns that have already been detected, ultimately reducing costs and the time it takes to complete certain tasks. Companies that want to harness the power of automated data must have data systems that are always online.

 

“Reaction is good, but preparedness and proactive decision making in an always-on world helps businesses excel.”

 

What’s your role in this trend?: “Veeam helps companies move to automated systems by providing reliable data management, including consistent backups so companies never have to worry about system downtime — which would otherwise be a huge expense due to lost data and productivity. Additionally, if companies have proper backups they can revisit historical data patterns to then predict future patterns and start to move data ahead of time. Reaction is good, but preparedness and proactive decision making in an always-on world helps businesses excel. Veeam has leveraged our customer support team’s large Hadoop big data instance where we would run machine learning algorithms internally to detect the root cause of errors. In doing so, we can quickly pinpoint the corrective action and build this into our product so that issues can be proactively identified and remediated.”

Great Companies Need Great People. That's Where We Come In.

Recruit With Us