Artificial intelligence is set to replace 1.8 million jobs by 2020.
That’s according to a recent Gartner study, which also estimates the AI industry will create 2.3 million new jobs by the same year — for a net gain.
The field’s hot, and there are thousands of openings for machine learning engineers, data scientists, research scientists and business intelligence engineers across the board. Interested in getting in on the action? We asked some local AI experts to weigh in what it's like to work in the burgeoning industry.
CognitiveScale’s director of product Amanda de la Motte said it takes a visionary with a futuristic sensibility — someone who also exudes both patience and impatience — to thrive in the field. She has held previous product leadership roles at Orckestra and RetailMeNot and said the biggest difference in working in AI is in the actual development of the product.
How do you explain in layman’s terms what CognitiveScale does?
We build industry-specific augmented intelligence solutions for financial services, healthcare and digital commerce markets that emulate and extend human cognitive functions by pairing people and machines. Our augmented intelligence solutions drive customer engagement and conversion, improve decision making and empower employees.
What’s the biggest challenge your team faces?
I'd say the biggest challenge we face is placing the right strategic bet. With finite resources across several highly-specialized disciplines, it is critical that AI startups stay focused and not try to boil the ocean.
Ours has been a challenging and exciting journey iterating over the various directions the market has taken us, but we are as confident as ever that our approach of building a bridge between data scientists and app developers to democratize AI for the enterprise is the one that will lead us to victory.
Compared to other tech verticals, what’s different about working in AI?
The largest difference is the experience of building the product. In other tech verticals, products are discrete ensembles of code that can be utilized/sold repeatedly with little to no alterations or customization. They also may or may not have a UI component or require data flowing into them.
AI products are more like full stack systems composed of multiple layers such as data, intelligence/learning and experiences that all need to be configured based on the use case and what data is available to you. They require feedback for ongoing learning and need to be trained with nutritious data that's highly relevant to the problem you're solving.
This boils down to a completely new, quite challenging way of building a profitable product in that you must work to find the components that are repeatable to avoid building custom solutions with each instantiation. Additionally, working in AI requires a hyper-scale ability to handle changing market dynamics and requires SME-level knowledge of industry-specific needs, requirements, compliance and regulations to be able to identify the types of business outcomes that can best benefit from AI.
Founded in 2015, Digital Nebula provides AI enterprise services like consulting and product creation. Justin Turner, VP of engineering, said some of the products the Digital Nebula team has developed include autonomous commercial drones, help desk bots and document redaction software.
Can you explain in layman’s terms what Digital Nebula does?
We are a boutique artificial intelligence services firm that provides tactical implementations of artificially intelligent solutions. We help customers identify AI use cases that reduce costs and increase efficiency. Then we provide the resources needed to execute the idea.
We also have a research and development side of our business where we push the envelope of artificial intelligence.
What's the biggest challenge your team faces?
Many companies are just getting into the area of artificial intelligence, and they may have heard about AI but may not necessarily know what it is. A large part of our business involves educating our clients about strategic AI solutions. That education mentality is required from everybody, from our sales team down to our engineering team.
Compared to other tech verticals, what's different about working in AI?
We are on the cutting edge of software engineering. It's an amazing and profound feeling when our engineers are able to make software which learns and makes better decisions than humans.
Cerebri AI is the second science-driven software company Jean Belanger has built from the ground up. Belanger talked with us about how the startup’s patent-pending technology analyzes data points from a customer’s journey to identify which specific events led to a purchase.
How do you explain in layman’s terms what Cerebri does?
We work with Fortune 500 companies to help them anticipate what their customers want. Companies must now be proactive rather than reactive as a result of the advent of the internet and popularity of mobile devices. A customer is no longer an address or a phone number; rather, a customer is now a customer journey: a series of events comprising their interactions with a supplier.
This sounds straightforward: gather up your customer data and put it in some semblance of order; however, it is not that easy. This is where AI and machine learning come in.
The problem AI solves is that the first event informs the second and the first and second events inform the third event. This knock-on effect, or the cumulative experience of a customer building up to a purchase, is almost impossible to solve without AI and machine learning.
The power of AI and cloud computing means Fortune 500 companies perform hundreds of millions of calculations to arrive at surprisingly subtle understandings of what events contribute to purchases, etc. This is the problem Cerebri AI solves with patent-pending technology we are developing here in Austin and in our lab in Toronto.
What’s the biggest challenge your team faces?
Our biggest challenge is to keep up with the interest in AI. This is the first technical discontinuity that is truly global in nature. Companies around the world are building and buying solutions while vendors from around the world are attacking the problem.
AI and machine learning generate decision options you simply cannot get any other way. Large enterprises are pushing AI technology forward at a blistering pace. Exciting stuff.
Compared to other tech verticals, what’s different about working in AI?
Science-based software is just plain hard to do. This is the second science-based software company I have helped start from a blank sheet of paper, and it is proving as challenging as ever. Any solution that is essentially math-based requires tremendous rigor and discipline to bring to market. But the excitement of discovering new ways to do things is totally awesome, and the days fly by.
Dan Greff, the VP of engineering and co-founder of Olono, said new approaches, algorithms and techniques constantly emerge because of how young the space is. If interested in entering the field, he said candidates must remain curious about the evolving industry — and the problems they’re trying to solve.
How do you explain in layman’s terms what Olono does?
Olono acts as a GPS for sales professionals — we provide intelligent, data-driven hints throughout complex B2B sales journeys. With data about prospects scattered throughout dozens of systems in a company, we use AI and machine learning to deliver Next Best Actions that help sales reps break through the noise and understand the most effective steps to close deals faster.
What’s the biggest challenge your team faces?
Recruiting and retaining experienced developers is a challenge across the tech industry, and the AI field is no different. We're fortunate to be in a place where our developers know they are leading a real transformation within the sales industry and are excited about the challenges they are solving. Beyond hiring, it's important not to get caught up in the hype and buzz surrounding AI and to stay focused on the data and the product priorities.
Compared to other tech verticals, what’s different about working in AI?
In some ways, it's not all that different. However, AI requires a massive amount of actual data. You can't just create software and expect good results without real datasets. That means the best results come when developers and data scientists work together from the very start of any AI or machine-learning project.
