Recently, Keyvan Mohajer, Co-Founder and CEO, SoundHound Inc., talked with Pejman Nozad, Founding Managing Partner, Pear VC, an early seed technology fund in Silicon Valley, as part of a video series focusing on successful Iranian-American founders. As a co-founder of SoundHound Inc., Keyvan has taken the vision of a voice-enabled world and made it a reality with the Houndify Voice AI platform.

During the discussion, Keyvan talked about the early days at SoundHound Inc., it’s founding, the lessons he’s learned over the past 20 years spent scaling multiple companies, and a few secrets to his success. You can listen to the interview in its entirety here, or read the following recap to get the highlights.

Founding SoundHound Inc.

Pejman: Tell us about SoundHound and why you started it. 

Keyvan: We are a 15 year old startup with the vision to build voice AI as a platform. We specialize in speech recognition and natural language understanding—enabling any company that has a product to add voice AI to the product. We think voice AI is the future of human and computer interaction. 

We started SoundHound as a voice AI company, that was always our vision. Fifteen or 20 years ago, we decided that we needed our own tech to be a major player and make a big impact as a disruptor—both in terms of technology and business model.

It turns out that building the voice AI takes 10 plus years. It’s not practical for a startup to spend 10 years in R&D, so we launched several products out of that research.

One of them was a music identification product that understood both recorded audio that’s playing in the background, and if you sing or hum a song with your voice. The SoundHound app turned out to be very successful. It has been downloaded more than 300 million times and it basically funded the first 10 years of our operation.

During that 10 years, the majority of the company was focused on building the Houndify Voice AI platform, which was really our main vision. At the time, it was the best kept secret in Silicon Valley. 

The entrepreneurial spirit

Pejman: When you started SoundHound you were still a PhD student at Stanford University studying electronics.

How did you start SoundHound? 

Keyvan: I had some entrepreneurial experience. I actually started my first company when I was younger than 10 years old. 

There was a children’s playground in Iran where kids played every evening. I made a little bonfire by the playground and I bought one bread dough for one unit of money. Let’s say one dollar. And I turned that dough into 100 little mini-breads and I cooked them in front of the kids. 

The value I delivered was obviously enjoying freshly baked breads, but also experiencing performing a transaction—because children really like to be like grownups. It was easy for them to get one unit of money from their parents and they would bring that and exchange their money for my bread. 

Everyday, I turned one dollar into a hundred dollars. Then, I came up with the idea of upselling. I had added lemon flavor for two dollars and chocolate flavor for three dollars.

At a young age, I had this itch of creating businesses and creating value.

When my family moved to Canada in 1995, I finished high school and engineering at the University of Toronto. There, I started three companies. By the time I graduated, I had some experience and I decided that I wanted to be the technical founder of a high-tech startup that would make a big impact in the world. 

I didn’t really want to be a serial entrepreneur. I wanted to make a big impact and I wanted to be a technical founder.

With that decision, I needed to go to grad school.

That brought me to Stanford in the year 2000. I searched for what that big thing was. The question that I came up with was, “What would happen in my lifetime, that would make a big impact in the world?”

To answer that question, I turned to SciFi and started watching Star Trek. There are a few concepts in Star Trek that didn’t exist then. There were five of them:

• Traveling faster than the speed of light
• Teleportation
• Holodecks
• Replicators
• Voice AI

The one that I thought for sure would happen in my lifetime was voice AI. This was 20 years ago.

In the science fiction movies, people talk to robots, talk to computers, talk to rooms, and there were devices that talked back and had a conversation—it just seemed very normal.

I thought that should happen in my lifetime and I wanted to be the one to do it. 

How do you compete with giants?

Pejman: A question I ask founders who are building something in a market where giant companies have the same kind of technology or product:

How did that influence you and how did you face those challenges?

Keyvan: Even now people ask, “How do you compete with the giants?”

I think the first step is to not be afraid. If you have a passion for something and you’re committed to it, don’t be afraid. Even those big giants used to be small.

Make sure you hire people who are not afraid, and raise money from people who are not afraid. Once you have fearful people around the table, they can actually pollute the environment. 

We didn’t want to be an independent player that has any inferior technology. We wanted our technology to be better than the giants—and maybe that’s why it took 10 years. For the first six years we kept failing. But, we kept trying and finally in 2015, we knew that what we had was very special and unveiled it. Even after five years, those giants haven’t been able to catch up.

When we talk to each other, we talk with complex conversations, but when we talk to computers, we have learned to lower our expectations. Why is it that? The answer is: limitation in natural language understanding, and that’s what we really wanted to fix. 

Our vision was a technology that wouldn’t have any limits in how complex the conversation can get. If you ask Siri, Google, or other assistants, “Show me restaurants, excluding Chinese,” they all usually respond with. “Here are some Chinese restaurants,” because they matched the keywords of Chinese and restaurants. With our technology, you can be a lot more complicated.

Here’s a demo:

So I was able to ask very complex questions, exclusions, double negation, compound criteria, and it even read it back to me to give me confidence that I’m not just getting keyword results, I’m getting exactly what I’ve asked for. I can also follow-up and refine my request. 

Stay the course—even if it takes 10 years

Pejman: How do you balance the day-to-day execution with keeping focused on your long-term vision?

Keyvan: You have to constantly zoom in and zoom out. Zoom in to execute and then zoom out to make sure you’re on the right path. You have to constantly grow with the new challenges. For us, one of the biggest challenges was the time it took. When something takes a long time, there are people who have doubts.

I had to constantly re-inspire people and make them remember that we believe in the individual and to not deviate from the goal that we had set. That was very challenging. 

Pejman: You are a leader of the multi-billion dollar company with 400 employees. Your customers are large enterprises, like Mercedes-Benz.

How did you transition from being a technologist to a CEO running such a massive operation?

Keyvan: I think constantly growing is the key, making sure you acknowledge the areas you are not good at, becoming good at those, and constantly pushing your boundaries.

You have to surround yourself with people who are better than you in what they do. Everybody around me is better than me in what they do. I think the biggest advice that I learned on my own was don’t follow advice that you don’t understand.

That’s one of the mistakes I made in the early days. Culturally, I’m very polite and really respect people who are older than me. But, sometimes people who are older than me, wiser than me, and more experienced than me told me to do this or do that.

That politeness in me kicked in and I did what they said without really understanding it, without believing in it. Some of that advice was good, but some turned out to be wrong for us.

Love what you do

Pejman: What are some good habits you try to keep in order to help preserve your motivation and drive for success?

Keyvan: You have to enjoy your work because it will be difficult. Success is not easy. So, if something’s going to take a long time and it’s going to be very difficult, you have to enjoy it.

If you wake up one day and you realize that you’re not enjoying this or that somehow you’ve gotten on a trajectory that is wrong for you, you need to make sure you fix those.

If you don’t enjoy it, you won’t last. 

Pejman: What does a typical day look like for you?

Keyvan: I used to not be a morning person when I was in college, but I have become a morning person. I wake up really early, I like to have a few hours before the work begins and that’s the time where I really do my magic because there’s no distraction. I’m on top of the world and I can really innovate and make sure I make the best decisions. Then, the rest of the day is just meetings, time to exercise, and making time for family.

Don’t just be a good CEO, executive, or technologist. Be a good husband, a good father, a good friend. Keep it balanced. 

The future of voice AI

Pejman: What will our world look like in 20 years?

Keyvan: I think the other things that you see in science fiction will happen. I think we will have autonomous cars, and voice AI will be a big part of our lives. People talk about 70 billion IOT devices. These devices will be too small to have a keyboard and a mouse, but they can always have a small microphone. You will have all these things that are connected, that have a microphone, they can hear you and you can hear them and talk to them.

I think there will be multiple assistants. Companies like Google and Amazon, they are trying to have one assistant like Google everywhere or Alexa everywhere.

We think that there will be multiple assistants.

Imagine in the future we live among robots. So 10 billion people and 20 billion robots. Will all those robots have the same name? Will they all be called Google or Amazon or Alexa? Obviously not every robot would have its own name, but they will have their own personalities and they will be good at different things.

Some of them will be doctors, teachers, cleaners, and some have to be your friends. You’ll learn to know their names and their personalities, and you’ll learn to respect them and coexist with them. That’s the future.

Leading with courage 

Pejman: Do you think being an immigrant and being part of the culture that values science and math has shaped your love for technology and being an entrepreneur? 

Keyvan: I have been a SciFi fan most of my life. It really inspires me and gives me a lot of motivation to make science fiction a reality. 

Growing up, I did well in math and I did well in school, but that’s not enough to do well as an entrepreneur. You need to have multiple dimensions. I’ve seen people who are smart and creative, but didn’t have the courage to do things that others won’t. You need to have stamina because it takes a long time and you also need to have some stars aligned for you. I call those the ingredients of success.

You can learn to be smart and creative, you can push yourself to have courage and to have stamina. You can achieve luck by making sure you have time—and you buy time by raising a lot of money.

You can hear the rest of the interview, including some personal insights from Keyvan here. Learn more about Houndify’s independent voice AI platform at Houndify.com and register for a free account. Want to learn more? Talk to us about how we can help you bring your voice strategy to life.