Insights From a Creator of Salesforce Einstein

I recently went one on one with Arijit Sengupta, a highly accomplished entrepreneur who sold his last startup, BeyondCore, for $110 million to Salesforce to expedite their creation of the AI platform that is now known as Salesforce Einstein. Arijit is also the founder and CEO of Aible, a software company that enables business people to create real world AI that delivers measurable business impact.

Arijit held leadership positions at several big data, cloud computing, and e-business industry associations, and previously worked at Salesforce, Oracle, Microsoft, and Yankee Group. He has been granted seventeen patents. Arijit has guest lectured at Stanford and other universities; spoken at conferences in a dozen countries; and has been written about in The World Is Flat 3.0, New York Times, San Jose Mercury News, Harvard Business Review, The Economist, and other leading publications. Arijit holds an MBA with Distinction from the Harvard Business School and Bachelor degrees with Distinction in Computer Science and Economics from Stanford University.

We spoke about entrepreneurship, leadership and AI:

Adam: Thanks again for taking the time to share your insights and advice. First things first, though, I am sure readers would love to learn more about you. How did you get here? What experiences, failures, setbacks or challenges have been most instrumental to your growth as an entrepreneur and a leader?

Arijit: I’m the founder and CEO of Aible. Prior to that, I founded a company called BeyondCore that was sold to Salesforce and has become part of Salesforce Einstein. I have more than 20 years of experience working with AI and have worked on more than 1,000 AI projects. I also wrote a book titled AI Is a Waste of Money, which describes why most AI projects fail and how businesses can create successful AI.

It’s been really interesting to see how hard it is to create something different and take it to market. It’s a question of first making sure your technology works. You have to be sure that you’re solving a true customer problem, not what you think the customer problem is. You also have to figure out how the customer will buy. Just because you have a solution doesn’t mean a customer can find it and buy it. And of course, getting the right team together is essential. You have to ensure you have different strengths in the organization, and this can be really difficult to do as an entrepreneur.

Adam: In your experience, what are the defining qualities of an effective leader? How can leaders and aspiring leaders take their leadership skills to the next level?

Arijit: The defining quality of an effective leader is your impact on the people around you. Leadership is all about being a catalyst. A true leader can actually change the nature of the people around them and empower them to achieve more than they think they can. A really important part of being a catalyst is constantly learning. Many times, the place where you think your strengths are is where you’ll find your biggest blind spots. You have to have people around you who will honestly stand up to you and push back when they think you’re wrong. 

As a leader, one of the most difficult things to do is bringing out the best in employees without pressuring them to a point where they break. It’s a difficult balancing act. You want to build an organization where the employees are motivated to deliver on the mission of the company rather than because they’re constantly being told to go faster. It’s about empowering your team. I usually set some pretty high expectations, but I encourage the team to know that they can do it.  Nine times out of ten, they not only live up to my expectations, they exceed them. They surprise me. And they surprise themselves.  

Adam: In your experience, what are the defining qualities of a great entrepreneur? How can entrepreneurs find the right idea, scale it, and as you did, position themselves for a big exit? 

Arijit: I always liked the Harvard Business School’s definition of entrepreneurship as the pursuit of opportunities beyond resources currently controlled. That means you have to think of opportunities in a different way.  For me, the defining quality of a great entrepreneur is looking at a problem and saying, “How can I change the way the market has approached this problem?” That way, you create a real step-change rather than just an incremental change.

In AI, everyone was training models to maximize academic metrics like accuracy, rather than to deliver what businesspeople really want from AI – measurable business impact. So instead of accuracy, we’ve focused on getting users directly to business impact. That’s a very different approach than what the market is doing. 

Adam: What are your three best tips applicable to early-stage entrepreneurs? 

Arijit: One, make sure that the customer will buy. Two, make sure you have a way to create value that the customer immediately recognizes. And three, make sure that you build a team that can deliver a solution in a trusted way. Way too many entrepreneurs get stuck on their great idea and fail to follow through. If you can’t deliver on the vision, the idea doesn’t matter. 

Adam: What are your three best tips applicable to executives and business leaders? 

Arijit: First, don’t tell people what to do. Explain to people where you’re trying to go and why you’re trying to go there. If you make that clear, people will surprise you with new ways to achieve the goals. If you’re simply telling people what to do, the fundamental scaling constraint is your ability to correctly come up with what to do every time. Management is not about telling people how to do things; it’s about telling people what needs to be done. They’ll often come up with better ways of doing things than you could have thought of because they’re on the front lines. 

Second, you not only have to hire the best people, but you also have to prune out the people who turn out to be wrong for the organization. I believe in having a very politics-free environment. People should be evaluated on the quality of the work they do and the content of their character. Everything else is just a distraction.

Third, you have to have a very clear idea of your company’s core values. At a bare minimum, that means making sure you’re not hurting the user or exposing them to unintended consequences. We see this a lot with traditional AI that’s maximized for accuracy because that kind of AI often winds up being very conservative in its predictions. In the pursuit of accuracy, the AI may tell you to go after only a few sure-fire sales. But that can be very bad for business if you miss out on a lot of deals you could have closed. Business impact beats accuracy every time.

Adam: You have strong views on AI. What do you believe are the key mistakes most executives make when deploying AI? How can businesses best extract value from AI? 

The way business can best extract value from AI is this: start from the end. Traditional AI evaluates many models based on accuracy and then leaves it up to the businessperson to choose the right model. The better way to do it is to reverse the process. First, decide exactly what your business objective is and then make the AI fit your objective. You figure out what you’re trying to achieve, what the benefits of getting a prediction right are, and what the costs of getting it wrong are. You factor in your business resource constraints, such a the size of your sales force or your marketing budget. Only then do you bring in the AI to optimize for real business impact. 

You hear a lot of business users say that their AI is never wrong – it’s just completely useless for what they’re trying to accomplish. That’s because most AI is trained to optimize for accuracy – getting most predictions right –  rather than business impact. An AI optimized for accuracy gives equal weight to the cost of failing to predict something as it does incorrectly predicting something will happen. But in business, the benefit of a correct AI prediction is never equal to the cost of a wrong one.

For example, what if the benefit of winning a deal is 100 times the cost of pursuing a deal? In that case, you might be willing to pursue and lose 99 deals for a single win. An AI that finds only 1 win in 100 would be very inaccurate. But it would boost your net revenue. The most accurate model is often not the best in terms of business impact. In fact, a perfectly accurate AI can run a company right out of business. 

Adam: How do you envision AI impacting our personal lives and businesses over the next decade? What are the key factors companies should consider when investing in and deploying AI? 

Arijit: Fundamentally, AI is going to be far more transformative than computing or the Internet has been. That’s because AI is instantly at scale. You can take a powerful predictive model, deploy it to your enterprise applications, and immediately affect the behavior of all of your users. That’s never been possible before at such immediate scale.

That’s AI’s strength, but it’s also its biggest weakness. One of the concerns I have is – do we want to become passive consumers of AI? Or do we want to have AI that gives us a say in how the AI is trained, that gives us some agency? If one salesperson prefers to pursue a few leads but do them very well, and another salesperson likes to pursue many leads and churn through them very quickly, why should the two of them have the same AI? We need to give people a say in their experience of AI, not just in their consumption of it.  

Adam: What is the single best piece of advice you have ever received? 

Arijit: Don’t believe people when they tell you something is impossible.

Adam: What is one thing everyone should be doing to pay it forward? 

Arijit: Everyone has to find their own way of making a difference. For me, I believe that if AI continues to go down the path it’s going, we’ll be in a world where a small set of people will create AI that will tell us everything we should be doing – who to meet for lunch, what articles to read, what clothes to buy. I don’t want that world for my 3-year-old daughter. I’m doing what I can to prevent that kind of AI world from coming to pass. When my daughter grows up, I want to be able to look her in the eye and say, “I did my best.”


Adam Mendler is the CEO of The Veloz Group, where he co-founded and oversees ventures across a wide variety of industries. Adam is also the creator and host of the business and leadership podcast Thirty Minute Mentors, where he goes one on one with America's most successful people - Fortune 500 CEOs, founders of household name companies, Hall of Fame and Olympic gold medal winning athletes, political and military leaders - for intimate half-hour conversations each week. Adam has written extensively on leadership, management, entrepreneurship, marketing and sales, having authored over 70 articles published in major media outlets including Forbes, Inc. and HuffPost, and has conducted more than 500 one on one interviews with America’s top leaders through his collective media projects. A top leadership speaker, Adam draws upon his insights building and leading businesses and interviewing hundreds of America's top leaders as a top keynote speaker to businesses, universities and non-profit organizations.

Follow Adam on Instagram and Twitter at @adammendler and listen and subscribe to Thirty Minute Mentors on your favorite podcasting app.

Check out https://www.salesforce.com/crm/crm-for-small-business/ for more information on CRMs.