Why a 22-year-old skipped college to build an AI company

This talk-show article is based on an interview with Ishraq Khan, the 22-year-old CEO of Kodezi in the San Francisco Bay Area. Updated for length and clarity.

I moved to the US from Bangladesh with my family in 2011. My path to becoming an AI company founder started with my interest in programming.

When I moved, my dad bought me a laptop, and I stumbled upon YouTube. One video I watched was a coding tutorial, which made me start learning coding and falling in love with it.

During my freshman year of high school, I was in computer science, and I noticed that students were spending too much time editing code in class. I was surprised, why isn’t Grammarly for developers automatically correcting coding errors?

I spent a lot of time trying to figure out machine learning, how we can make code improve itself, and how to automate the whole process. It took me about a year, but I got myself to work in the model.

This was the beginning of my company that I am working full time at 22.

By my senior year, I got my first VC call

A VC approached me and said they liked what I was building, but there was no money to be made, so I realized that I should focus on something even bigger, which became my company, Kodezi.

Starting an AI company at 17, the biggest step I took was writing emails. I found out that you can send an email to anyone by Googling their email address. I emailed CEOs, startups, AI researchers, and venture capitalists.

I reached out to try to get internships and learn from people who were ahead of me. Over time, those emails turned into conversations about what I was building with some really big names.

I found an event called Orlando Synapse, and I sent an email and said, “I’m a senior in high school, and I don’t have $500 for a box. Is there any way I can come up with it for free? This is what I’m building.” Someone answered within a few hours and said, “Sure, here’s a free house.”

I posted on LinkedIn that I went to this event and found one of my first angel investors

I earned $20,000 before I turned 18.

My biggest challenge was getting others to say yes from an investment perspective, because back then, there was no ChatGPT, and AI applications were still considered experimental.

It was difficult to convince investors to support an AI-first code platform from a young age. There was doubt not only about the technique, but about whether I could kill at that level. Once developing AI became more widely accepted, the narrative around which we were building became easier to understand.

I raised $800,000 before I turned 19 and $2 million before I turned 22

One of my ways to do this was to learn to answer three questions from investors:

1. What are you building?
2. Why does this matter?
3. Why would this be a billion-dollar company?

I also had to think about how to make them see that I am the right person to do it.

I overcame interviewer rejection by making calls while interviewing them, rather than being interviewed. I started trying to understand who they are, why they invest in these companies, and what their purpose is. This helped me decide if they were worth it.

Another challenge was deciding whether to go to college or work full time at my company when I was 17

I applied to 60 colleges, got into more than a dozen, including Ivy League schools, and finally decided to drop out of college altogether.

“I want to go to college” was still the answer I would give people, and when salespeople asked, I would say, “I don’t really know, maybe…” My uncertainty about it was another reason why other salespeople were unsure.

I realized that I would hate it if I went to college, because I wouldn’t have the same opportunities, and I could always go to college later. If I wanted to build an AI company after I graduated, thousands of others would have already done the same.

Many of my friends had graduated from college and gotten jobs, but this was my job now

I don’t regret dropping out of college. I think college offers a built-in community environment that is hard to emulate. It’s easier to make friends when you’re surrounded by your peers at a different stage of life.

That said, the people I spend time with right now are builders – small startups, owners, users, and ambitious executives who want to create something meaningful. That sense of sharing is powerful.

There is always an opportunity cost, and for me, that trade-off was important.

For others who want to start an AI company, I recommend doing whatever you are doing as soon as possible

Without feedback from people who aren’t your friends, you won’t really know if what you’re building is connecting with people.

Also, don’t aim for a metric like “success” that will stop you. Instead of chasing external success, chase internal abilities, because that will lead to successful results.

I’m 30 years old, and I see that I’m still building my company

I think the world needs more innovation, and that’s where young entrepreneurs start, especially with the rise of AI tools.

I have been leading the company for six years. We have grown to over 35 employees. My next goal is to manage the code security layer of enterprise companies.

I want Code to be the unique system companies rely on to keep their codebases healthy over time. If writing code is like building a car, then maintaining it requires mechanics. Our mission is to be an automated software engineer.