Math: AI is math with a well-designed cover, and to shape the technology’s future, it’s an essential skill. GenAI is about producing the right probabilities, but it doesn’t understand what you’re saying. It is doing mathematical calculations in order to produce the poem, piece of code or image you requested. The current architecture presents limitations because of the math being used. Having some math familiarity can help identify where AI can improve.
Artificial Intelligence is transforming industries at a breakneck pace, and the entrepreneurs driving this innovation are at the forefront of this revolution. From groundbreaking applications to ethical considerations, these visionaries are shaping the future of AI. What does it take to innovate in such a rapidly evolving field, and how are these entrepreneurs using AI to solve real-world problems? As a part of this series, I had the pleasure of interviewing Eduardo Mota.
Eduardo Mota is senior cloud data architect — AI/ML specialist at DoiT. An accomplished cloud architect and machine learning specialist, he holds a Bachelor of Business Administration and multiple Machine Learning certifications, demonstrating his relentless pursuit of knowledge. Eduardo’s journey includes pivotal roles at DoiT and AWS, where his expertise in AWS and GCP cloud architecture and optimization strategies significantly impacted operational efficiency and cost savings for multiple organizations.
Thank you so much for joining us in this interview series. Before we dive in, our readers would love to learn a bit more about you. Can you share the most interesting story that happened to you since you began your career?
I was hired by an organization to help them modernize their contact center. The first task was to improvise the interactive voice response (IVR) phone system. There was no documentation, and changes had been made over a period of years by many people. At the end of my first week, I sat in front of the configuration page making the last change to my IVR diagram. I moved my notebook and hit the mouse by accident, which caused the page to refresh. I checked and everything seemed to be ok. Everyone was on their way out of the office when I decided to check the phone system to ensure everything was working properly. I dialed one of our numbers and got the message: “The number you have dialed is not in service.” I dialed two other numbers and got the same message.
I sat there for a second trying to figure out what I had done. Without answers, I turned around and went into my boss’s office, her name is Gina. I told her what had happened, and she simply asked, “How can I help?” I asked for the phone number of our vendor managing the phone system. After a couple of rings, an engineer answered the call. I explained the state of the phone system, and15 mins later, the system was restored. I had deleted an entry that connected the phone number to the IVR system. The tool we were using didn’t use confirmation popups or confirmation of changes. I let Gina know everything was fine and off I went.
The next day, I saw the owner of the company go into Gina’s office. After 15 minutes, Gina called me in, and every scenario of how I was going to be fired went through my mind. The owner in a very serious tone asked: “I received the report you sent yesterday, are the numbers correct?” I replied yes. He then said, “Ok, that’s everything, thank you.”
I never heard about the IVR incident again. A few years later, I went on vacation and during that time a contractor joined the organization. She changed an account setting when she was logged in to the IVR system, thinking it was a local setting and not a global one. Documentation was also poor. By the time I had gone back to work, she was no longer with the company.
I always remember this story because a mistake is rarely a one-person event. There are managers and leaders to help, and my AI journey in particular wouldn’t be possible without them. Equally important were the mistakes I made and chances I was given to rectify and learn from them. That moment for me changed my approach to risk management and how I evaluate risk with my teams.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
There are two. The first is Dale Moury who at the beginning of my career taught me how to be a great leader. He knew how to build trust and helped me push for what I wanted. The highlight of his approach came when I had an opportunity for a new role and he helped me make the best career decision even when it meant leaving his team.
The second person was Carlos Gil. He pushed me to develop my technical skills further and explore different ones, too. He encouraged my passion for AI/ML which has led to what I’m doing today. And not only did I learn the technical skills, but I also had opportunities to influence others through such things as presentations, supported by a strong belief in the value of what I was delivering.
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
I live by the quote: “Difficult things take me time to do, and impossible things take me just a little longer.” Everything is possible and limitations are self-imposed. I address every problem with the attitude that there is a solution. People are more receptive to work with you, and be encouraging, when they truly see you are trying your best to create a positive outcome.
I started my career in customer services, moved to project management, then DevOps, and finally data management. In my personal life, I moved from Mexico to Canada at the age of 16. These changes were not easy, there were a lot of unknowns ahead, and in many instances there was no clear path. This mantra of being able to do anything — regardless of how difficult the situation — I find comforting. There is no other way for me. In my career, every problem has a solution, it may be complex or may take time, but there is always a way to do something.
You are a successful business leader. Which three character traits do you think were most instrumental to your success? Can you please share a story or example for each?
Passion — I started learning ML and data processing before it was cool. The passion to do what you love takes you into adventures you never imagined. You can find yourself on a magical path where everything falls into place and just works. That’s been my journey with ML.
Humanity — Trust, respect and communication are the three main characteristics of humanity. I remember a particular situation a long time ago, in which the organization was restructuring and there were difficult decisions to be made. In my team, I had the person that I relied the most on, and needed her help when this restructuring would unfold. The mandate was to not share any info on the changes until an official announcement was made. I couldn’t do that to this person and decided to be human. I respected her efforts and opinion, trusted her to keep the changes to herself, and decided to tell her what was coming.
It was one of the best decisions I made. Not only did she have time to process the changes, she was able to prepare a plan to deal with all the system changes required. After this situation, she was further committed to the team and getting us through the changes, d while helping me bring morale up again in the department.
Risk tolerant — I wouldn’t necessarily call myself a risk seeker, but to be successful, you must be comfortable with risks. I have taken to heart a great lesson I got while working at Amazon: Make fast decisions when the decision can be reversed. Now working at DoiT, I get a chance to constantly experiment and move fast. This has allowed me to have success, but there is also a pile of lessons I learned when things didn’t work out. I refuse to call them failures as those were the steppingstones to building something great.
That said, test the boundaries of what technology can do, and if it shows promise, go after it. One of the latest examples for me was a project where we needed to analyze a year’s worth of data and identify why it behaved the way it did automatically. The risk of failure was huge as this solution was for a huge organization. So I jumped in, figured my way around, found several ways of how not to tackle it, and then identified one that would solve the problem.
Ok super. Let’s now shift to the main part of our discussion. Share the story of what inspired you to start working with AI. Was there a particular problem or opportunity that motivated you?
During 9th grade I developed a particular interest in psychology. This led me into understanding how the mind works, which got mixed into my passion for computers and how they worked. Fast forward through many failed attempts to understand AI to2018. I was working in a FinTech organization and presented with a big headache. The business was growing and we couldn’t hire customer service agents fast enough, nor was it financially possible. Facing this challenge, I turned to NLP and statistics.
Through data analysis, I figured that 80% of emails received in the contact center were able to be solved by one reply. And the majority of the information was already available on the company’s website. So, working with my team, we built a system based on NLP models to identify the intent of an email and reply with a templated response. After three months, the system was handling 2500 emails a week which represented 70% of the emails received. This system further expanded into other workloads such as converting emails into phone calls.
The entire system was built in-house and my passion was solidified after seeing the impact of the work. Ever since, a question has remained in my mind: What else can be done with AI?
Describe a moment when AI achieved something you once thought impossible. What was the breakthrough, and how did it impact your approach going forward?
I think it was early in my career when I started doing analysis on data. We were forecasting our activity volume for the organization and creating models to be better prepared for the future. I was amazed at how ML was able to take into account many factors and give me a very accurate picture of the future. I was amazed because I thought human behavior was random. But I ended up finding that the larger the group of people the more predictable the behaviour is. Since then, I don’t get surprised by what technology can achieve — I know there is more to come and the next wave will be even more impactful.
Talk about a challenge you faced when working with AI. How did you overcome it, and what was the outcome?
A challenge is to explain how AI works to the business. Not only is it necessary to get buy-in for an AI project, but also to have a realistic conversation around the ROI potential. Early in my career, when I started using NLP, the most common question from the business and from customers was how does the ML system identify the intent of a question. Is it based on the number of worlds? Are you matching words to a dictionary? What’s happening here?
This led me to explore multiple ways to explain ML systems, sometimes leaving people more confused. Yet, this allowed me to create a vocabulary to bridge the gap between technology and business. Now, I start with the possibilities and what the outcome could be, and then work backwards to explain how the system accomplishes this. This allows the conversation to be anchored to tangible outcomes the business understands.
Can you share an example of how your work with AI has had a meaningful impact (on others, on business results, etc)? What was the situation, and what difference did it make?
Working with AI has made me realize the important of human beings being part of the loop. We do automation in general to help us achieve more, be more productive and create a higher value to our peers. This, by nature, means that we have to be part of that loop to ensure the value generated by AI is truly there.
This year I delivered several GenAI Accelerators to DoiT. This program allows you to get an MVP solution up and running in six weeks. Again, at the human level, I see the impact of using AI to help close the gap between ideas and results. Allowing developers to see the power of AI, and the possibilities it can open up in such a short period of time, makes it more tangible. I have seen first-hand how this sparks curiosity and exploration in teams. And it brings me a lot of satisfaction knowing that the work I do allows people to use AI in the most productive and responsible ways.
Based on your experience and success, can you please share “Five Things You Need To Know To Help Shape The Future of AI”?
There are many sides to AI and contributing to its future is not limited to technology itself. There is a huge ethical aspect of AI that still is at play. Depending on your interest in AI one or more of the following will be stronger than the others.
1. Data management: Properly controlling data is critical whenever working with AI. This part is important for security, processing and making the right data available to the right AI model. Over the years, organizations have accumulated vast amounts of data, a lot of it stored in relational databases. These are great at storing transactions, but these are not always made available to GenAI. How can all that data fit into the prompt? Do you even need all that data or will a calculation of the data yield the right answer? Data management is key to ensuring money is used efficiently, keeping data secure, and ultimately providing a better user experience when AI is involved.
2. Domain knowledge: This goes hand in hand with the previous point. Knowing your industry is critical for leveraging AI the right way, and understanding how it can evolve. Consider a salesperson who has had tremendous success selling cars. Transfer them to sales for a software company without any training and they’re going to fail or at best perform average. What’s required is to get them to understand how the industry operates and the nuances of selling software. This teaching should happen with AI when designing the data pipeline and making the right data available to the model. Knowledge of the industry is critical in order to give the model the proper data to make a decision, as well as for evaluating the quality of the model. This in turn allows us to identify and evaluate where the AI needs to improve and shape its development.
3. Customer obsessed: Individuals and organizations need to be customer obsessed to ensure the needs of targets are met. This is not necessarily new to AI, we see this with any business and with any technology. Today it seems every business has a customer service chatbot. The difference between an ok experience to a “wow” experience is when the chatbot can provide responses based on customer activity. Keeping the human in the loop and understanding how the customer experience is affected by AI is critical to ensuring it provides values and is used in a responsible and ethical way.
4. Coding: I’m not talking about being an expert in programming. This is about understanding how coding works, and having knowledge of any programming language helps. This is the road on which we drive AI on. If we do not know the limitations of the road, we will find difficulty driving our AI solutions on it. It’s simple to say that we need to connect two systems to produce a specific result. Talk to a developer and they will have a thousand questions about that statement. AI is no different, we need to know what our limitations are and how we can overcome them. The future of AI is not isolated to GenAI models, models need to be integrated with other systems, and as such we need to know the ecosystem to see how we can keep shaping AI in the future.
5. Math: AI is math with a well-designed cover, and to shape the technology’s future, it’s an essential skill. GenAI is about producing the right probabilities, but it doesn’t understand what you’re saying. It is doing mathematical calculations in order to produce the poem, piece of code or image you requested. The current architecture presents limitations because of the math being used. Having some math familiarity can help identify where AI can improve.
When you think about the future of AI, what excites you the most, and how do you see your work contributing to that future?
The AI journey is getting more exciting with every announcement. What’s most exciting to me is the possibilities of the unknown. Thirty years ago, we didn’t know this concept of cloud. Even as cloud technology emerged, it wasn’t clear everything that was possible with it. AI is following a similar journey at a much faster pace. We are entering an era where hyper-personalization is starting to take shape. There are going to be many innovations in the years to come that will allow us to express who we are as individuals and interact with products as services the way we want. It’s exciting to think that there is so much more to come.
Even so, we still need to ensure safety and ethical use of AI tools to create value. We cannot just talk about value from a financial perspective, or, without considering the value that a person receives before, during and after using AI. I see my work and the work of others in the industry to make this an important part of every AI solution.
What advice would you give to other entrepreneurs who want to innovate in AI? Can you share a story from your experience that illustrates your advice?
AI is a fascinating tool, but ultimately it is still a tool. This means that it won’t solve every problem efficiently on the first try. There is still work to be done, ad iterating through solutions is the way to efficiently integrate AI to an organization or product. Try, adjust and try again. AI has a lot of potential, and I don’t believe we have seen everything that current AI models can do. Avoid trying to get a perfect solution, have guardrails in place and expose AI to your users. Their feedback is going to be gold to maximize AI ROI.
For example, I was once tasked to create insights from time series data. The proposed solution was to expose an interface that would allow users to ask questions connected to the database. This was the first time for the team to work with AI and there was little experience integrating it to a product. These was a lot of emphasis on getting it right to generate value. After some discussion, we landed on a simpler approach to use predefined prompts to pull data and then get insights from this. The approach allowed us to reduce the security risk, test fast and get feedback from customers. Quickly we noticed the amount of data required to generate insights was much larger than anticipated and we needed to adjust the approach. We ended up with multi-step inference. The lesson here is that if we had gone all in to deliver the original desired output, we would have spent a lot of time solving security and user interface issues, rather than testing the fundamentals and iterating through it to create value right away.
Is there a person in the world, or in the US with whom you would like to have a private breakfast or lunch, and why? He or she might just see this, especially if we tag them. 🙂
Andrew Ng. The list of his accomplishments is long but being one of the founders and head of Google Brain, I’d love to hear him summarize the impact and importance of his work in the field. One of the first AI courses I took was from him and it inspired me to continue advocating for sage and ethical use of this powerful technology.
How can our readers further follow your work online?
Best way is https://www.linkedin.com/in/motaed/
Thank you so much for joining us. This was very inspirational, and we wish you continued success in your important work.
Eduardo Mota Of DoiT On Pushing the Boundaries of AI was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.