4 Min Read

AI in the Cloud: How Marketers Can Power Prediction, Personalization, and Performance [VIDEO]

Featured Image

Wondering how to get started with AI? Take our on-demand Piloting AI for Marketers Series.

Learn More

Amazon Web Services (AWS) powers much of the world—much of what you see, and much of which you don’t realize. And it has been doing so with artificial intelligence for more than 20 years. Long-time community member of Marketing AI Institute and Head of AI Solutions Marketing at AWS, Albert Esplugas, sat down with Paul Roetzer at MAICON 2022 to discuss technological advancements, AI, and what he sees as the future for marketers and businesses.

AWS, the world’s most comprehensive and broadly-adopted cloud platform provider, has more than 160 full-feature services. They have millions of customers using the cloud to advance innovations, scalability, flexibility, and cost reduction, all leading to increased revenue. AWS alone generated $19.7 billion in earnings in Q2 2022, making it a massive revenue driver for Amazon.

 

Kicking things off, Esplugas identified the five big transformational technologies of the last 20 to 25 years: 

  1. The PC — The personal computer’s impact on productivity and augmenting human ingenuity is very similar to AI’s impact today.
  2. The Internet — Suddenly, from any PC, you could access so many resources, providing so much intelligence across this connection.
  3. The Smartphone — This device provides mobility, building on the power of the internet.
  4. Cloud and cloud computing — The cloud was critical for transformational technology, allowing users to build on top of basic innovation. The cloud provides storage and allows companies the flexibility of being able to train their models and use the computing they need to train those models. Flexibility is a key success of the cloud for any company.
  5. Artificial intelligence — While the oldest of these technologies, dating back to 1950, the cloud has helped with the adoption of AI.

And businesses are on their way. Going from understanding to piloting to operationalizing and adoption across the organization, there comes that moment when the business value of AI is realized and deployed across the organization.

What Marketers Can Learn from AWS

Amazon always innovates based on customer needs.

Every engineering team within Amazon is working with top customers who are adopting technology, understanding their needs, and innovating and improving based on customer feedback.

Research is a core component of business growth.

Amazon also has Amazon Science, their research organization, which combines pure customer-centric innovation with the science coming from their research team. This drives the way they structure and evolve their products.

“You will use AI at some point. The earlier you use it, the stronger business impact it will have and the stronger competitive advantage it will give you.” ~ Albert Esplugas, AWS

Piloting AI and the Cloud

Organizations wanting to pilot AI first can work directly with companies like AWS, or can work with their partners within their large ecosystem. Or, they can work with independent software vendors (ISV) that are using AWS to provide their service

If you’re using AWS, there is a three-layer system:

  1. Develop your model, bring it to Amazon, and run it on their cloud to use their computing power and storage.
  2. Build an end-to-end solution and model using SageMaker, which ingests and cleans data, then trains and deploys the model.
  3. Use AWS’s AI service with pre-trained models.

AWS’s products include Amazon Comprehend (understanding), Amazon Forecast (prediction), Kendra (intelligence), and Lex (conversational), which are great for specific use cases and even specific to industries such as healthcare and industrial manufacturing.

Three Key Focus Areas of AI

According to Esplugas, successful organizations need to master three key areas: mindset, skillset, and toolset.

Mindset: You will deploy AI to solve a business problem. You don’t deploy AI just for the sake of deploying AI. You identify the business problem or the one thing where AI can have a strong impact. You need to know the business metrics that will be impacted, how you’ll measure success, and who will be involved.

Skillset: You need to upskill and train your team on AI/ML at all levels, from engineering to the business and operational side. You want to have a single conversation and understanding of what AI means for your business and society.

Toolset: Determine your tech stack, the services you can use, the models you can build, what it means for you, and how you will deploy these models within your organization.

That said, this won’t all happen at once. The initial impact may be low, especially if you only have one key focus area. You’re still learning. As your deployments increase, AI has the power to transform the organization completely. You can become an AI emergent company.

AI Can Help Achieve Your Most Critical Business Outcomes

Start to determine which business use case you can solve. AWS also launched AI Use Case Explorer, which might be a great place for you to start. AI business use cases include:

  • Reducing costs — The reason many companies start new projects (marketing AI application: automating backend systems, document processing).
  • Increasing revenue — Using AI/ML to have a better customer experience and stronger engagement, leading to stronger revenue (marketing AI application: personalization, forecasting, fraud detection, recommender engines).
  • Improving product quality — Using AI/ML to make sure your product/service is improving (marketing AI application: conversational AI, voice analysis, intelligent search).
  • Reducing risk — Making sure your brand is protected (marketing AI application: content moderation).
  • Driving efficiencies — Using AI/ML to drive efficiencies within the organization. (marketing AI application: natural language processing, prediction).

Pre-Trained Model or Build Your Own Model?

Do you need to hire data scientists? Do you need IT people? Do you use a pre-trained model? Do you work with AWS or someone in their ecosystem? How do you decide?

Esplugas shared the example of Intuit and their decisions in their AI-powered technology. Intuit uses AWS, and they have their core recommendation system based on a model they built. This is the core of their business, so it makes sense for them to have their own unique model. However, they have other functionalities added to their application where pre-trained models make more sense from a cost and time investment standpoint.

The commoditization of language generation

“Generative AI is a technology that is now exploding,” explained Esplugas. AWS recently launched Code Whisperer, a generative AI for coding. If you want to develop an application, you type the application, and as you’re typing, the system is capable of understanding your intention. It can correct your coding errors and tell you what you’re not doing right. This will increase the productivity of developers, not replace them. This is the same on the marketing side.

The goal of AI is to augment human ingenuity. ~ Albert Esplugas, AWS

Accelerating adoption for enterprises vs. small businesses

Business size could play a factor in the piloting of AI programs. Esplugas offered his thoughts:

  • Enterprise: Upskill the organization through training. For example, Amazon has Embark, a machine learning education program, for this purpose.
  • SMBs: Look at what other organizations are doing. What are the use cases, and the impact they have? How this will apply to your organization? Find partners with experience in your industry and segment.

Overall, find technology based not on whether they’re using AI but rather because AI is making that technology better.

As his final thought of the interview, Esplugas shared his excitement about AI with the audience, saying, “It's augmenting human ingenuity, improving productivity by using AI to serve humans. And that's a fantastic future that we have ahead, and I'm super excited to be part of it. And I think that everyone will be part of it.”

Become a next-gen marketer by checking out the resources at the Marketing AI Institute. Read our blog posts, take our Intro to AI for Marketers class, attend webinars, join our community, download reports, guides, and templates (all free), read Marketing Artificial Intelligence, look into AI Academy for Marketers and Piloting AI Bundle, and our annual MAICON—Marketing AI Conference.

Get access to all MAICON main stage keynotes, sessions, and panels with the MAICON 2022 On-Demand Bundle

Related Posts

Meet Karen Hao: Marketing AI Conference (MAICON) 2021 Speaker

Cathy McPhillips | July 28, 2021

Karen Hao is the senior AI editor at MIT Technology Review, and a speaker at the Marketing AI Conference (MAICON) 2021. Learn more about her work here.

Ethical and Trustworthy AI: Lessons from the Front Lines [Video]

Cathy McPhillips | September 26, 2022

Gemma Galdón Clavell’s keynote from MAICON 2022 around AI, ethics, and the practical work being done to make AI safer and more responsible.

How AI + Predictive Analytics Drive Marketing Strategy [Video]

Cathy McPhillips | October 4, 2022

Learn more about how AI + predictive analytics drive marketings strategy in Atalia Horenshtien’s MAICON 2022 keynote.