Stanford Startup Spotlight: Seer Tracking, Amber Yang

Stanford Startups October 24, 2018

Amber Yang is a sophomore at Stanford studying Physics and Computer Science who founded and currently operates Seer Tracking. Seer Tracking’s mission is to track and map space debris in Low Earth Orbit for the purposes of reducing space debris collision risk with space missions. Seer Tracking uses Artificial Neural Network software in order recognize and predict space debris paths and potential collisions.

1. Where did your company Seer Tracking come from?

Amber Yang: This actually started as a project that I did in high school, actually for a science fair. I got the inspiration for the project because I grew up in Orlando, really near the Kennedy Space Center, so I really loved space from a really young age and then, later on, when astronaut Scott Kelly spent his year in space I would watch his Facebook livestreams and a lot of times he would say that he would have duck out from the International Space Station and into another capsule because there was danger of the Space Station being hit by debris.

That really got me thinking — I saw a lot of articles in the Wall Street Journal and the New York Times talking about the space debris problem so I started reading some papers and realized that the current methods of tracking debris often times have a lot of error so people can’t really predict where the debris will go simply because there is so much drag and various atmospheric effects that are impacting the debris. These papers were also using a mathematical model that made it really hard to model drag and these other atmospheric and weather effects and so I had this idea to use artificial intelligence to recognize the patterns in how the debris’ orbits change over time to predict future orbits.

I had this idea when I was sophomore in high school and then I worked on it for about two years in high school, competed in some science fairs for it, and then after high school, I realized that with a lot of these science fairs, a lot of times the projects don’t really end up going anywhere and people end up abandoning these projects when they go to college and I thought that would be a big shame because these are some really big problems that need to be solved. I really wanted to see my idea come to use so that’s what gave me the motivation to create a startup.

2. What were the kinds of technical resources that you were looking at when you started your project?

Amber Yang: I actually programmed everything in MATLAB, which I know might make a lot of people at Stanford cringe, but MATLAB has an AI library that's actually really easy to use if you’ve never worked with AI before and I had never worked with AI before – actually, I barely even knew how to code when I started the project. So, MATLAB was a really easy language to pick up and then the AI library was very easy to implement from there which made my experience using AI a lot easier.

In terms of the data I used, at first, I actually simulated all of my data using a normal standard deviation to create random samples of debris but then I later realized that there are space debris databases publicly available online that anyone can access – websites like spacetrack.org. That’s when I started to do more testing with my idea.

I feel like when you have an idea, it can be really daunting to try implement it and a lot of times you might lack the technical skills, but even now, after I’ve finished developing a lot of my project, I still don’t think I’m fully versed in artificial intelligence, and I’ve only really tested my project on my personal computer — I haven’t looked into things like GPU’s.

When you are first starting out with your idea, I would try to make it as easy as possible as you can and then maybe once you’ve developed some type of prototype and you have the idea to create a company, that’s when you can really delegate harder, advanced technical tasks to people who are really well-versed in those things. So just really taking advantage of all your opportunities and resources and also asking for help — always being willing to learn and admit that you don’t know something.

3. What is the state of the company right now?

Amber Yang: The company is actually being incorporated next week! I also actually just got accepted into an accelerator based in Toronto, called Creative Destruction Lab, and their focus is actually in space startups so I’m actually going to be traveling to Canada once every two months to work on the company there and have mentorship from them.

Right now, I have not raised any funding for the company. I’ve bootstrapped everything because it's really been only me developing the software and I don’t need to pay myself a salary or anything and I’m trying to delay as much, as possible, the time when I give away equity of the company. I’m also actually currently talking with three potential customers, one of which I’m actually getting data from so I’m trying to test the robustness of the software and then also to trying to customize our first software packaging based on their experiences. Research and development and then creating a start up are very different — in order to operate all of the software right now, you really need to know how to code so I’m in the process of creating a user interface for it. We’re still a little bit in the development of the software stage but hopefully customizing it for the customer.

By the end of the year, I’m hoping to close the deals with these two customers and then potentially raise a seed funding round simply because I’m hoping to expand the team — right now it's just me, I have an advisor in the Aero / Astro department here and then a potential CTO — but its really hard being a student while also developing software while also trying to develop the business of the company. I’m definitely looking to expand the team and hire some technical people.

4. What has the development process been like for commercializing your software and making it commercially available?

Amber Yang: As I mentioned, everything’s written in MATLAB so my first plan of action has been to finish converting it to Python and C++ so that’s been a bit of a struggle to manage with classes. After that, in terms of software licensing, I have a lot to learn in that aspect so i’m really looking to bring on another person to the team who is experienced with software licensing, software development, and things like that.

A visualization of Seer Tracking’s space debris tracking software.

5. How are you balancing your schoolwork and schedule alongside your company work?

Amber Yang: It’s a lot! One thing that I’m hesitant to do is bring on new people especially because this idea started only with me, so it can feel weird to bring on other people, but it's a very necessary step. In terms of my day, I go to classes from 8:30 to 4 or 5 everyday and then work on psets at night but typically I try to work on code on Mondays, Wednesdays, and Fridays and then Tuesdays and Thursdays, I’m working with someone in the accelerator in order to come up with a business plan and scaling plan of how to scale the company. With a lot of these operations, I think it's dumb for me to hold all of it on to myself so right now I’m definitely looking to expand the team.

6. How has the process of learning how to start a company been for you?

Amber Yang: When I incorporated the company, I actually signed on with a lawyer firm here in Palo Alto and they’ve actually taken care of all the incorporation details for me and I think especially when you are a young founder, take advantage of the opportunity that you're young and that a lot of people are more willing to help you just because they know you have a lot in your future.

Honestly, when you are trying to learn the basics of starting a company, making connections is probably the single most important thing you can do, not only with lawyers or VC’s, but actually with other young founders. For example, I got to meet a lot of people from the Thiel Fellowship group and I think talking to them and asking them questions about the founding process has been really helpful for me.

At this point, when I have a question or a concern, either asking someone or looking it up online and learning as you go through it has been my process. I think that's why, when you look at a lot of arguments about why founders should drop out, you realize that even if I were to go to business school and get an MBA degree, I wouldn’t be in the nitty-gritty works of actually starting a company and getting it off the ground because I think that's where you can actually learn a lot of stuff from.
7. One might think that tracking space debris should be the government’s job or responsibility — why should a private company be handling this?

Amber Yang: This has been a problem that the government has been dealing with essentially since the space race and I would say that increasingly, we are seeing space getting farther away from the government in a way and I think that a lot of times government technology development can be very slow and I think that the reason that we are seeing a lot of progress in space recently is because a lot of these technologies are being handled by private companies — SpaceX, for instance, has developed so much in the past couple of years and I really think its because the energy of startups is so much different from the government. People are very excited to do new things and I think that because of that you can see real tangible things and technologies come to fruition a lot quicker. I would say, for example, the government knows about what I’m doing but I don’t think they’ve been exploring the idea of using AI to track debris, I think they are sticking with their method. The importance for having a startup to handle these technologies is to keep coming up with these new ideas and then hopefully dealing with other private space industries just so we have our own type of ecosystem that keeps pushing out its own new developments.

8. How are you thinking about your company future for the next 6 months?

Amber Yang: My biggest goal is to finish development of the software so that I can demo it as a finalized product — that is my first and foremost priority. I think that a lot of times especially in Silicon Valley, you can get caught up between trying to raise money and getting customers before the project is completely finished and a lot of times people end up releasing bad beta versions but I think that for me, I really want to focus on getting the software and the technology to the best state possible just because I’m dealing in a very risky business –– predicting where things go into space is quite a big deal so I really want to finish development, make the software very professional and then probably within the next 6 months, hopefully get some customers to agree to buy a license after the development is completely finished. Between that time and now, I also plan bring on some people onto the team and potentially raise a small seed round.