Being midway through my journey with Fellowship.ai, I decided to share some aspects that might help those who are considering applying to the program.

I first came across Fellowship.ai about a couple of years back when I was going through fast.ai courses. It sounded really compelling, but it was a big commitment for me at the time as the program was in-person and at a handful of cities. This meant that I had to relocate to one of those cities for the entire period of the program.

Luckily, this has changed and the program is now fully remote. This makes it a very convenient option for aspiring data scientists and machine learning engineers.

What is Fellowship.ai?

Fellowship.ai is a highly competitive program that provides you with hands-on experience working on real-world projects spanning across various tasks in an environment that resembles actual industry setting.

What do you gain by joining the fellowship?

In addition to the hands-on experience, you get to practice agile development methodology, work with a team of fellows with diverse backgrounds, and access to a network of experienced mentors. As an added bonus, you get to participate in the AI reading group where you present and get to know about cutting-edge research papers. Moreover, externships are typically offered by Launchpad.AI to top performing fellows.

What experience is required?

As mentioned earlier, you will be working on implementing leading-edge solutions to solid business or research problems. Hence, you are expected to be well-versed in writing code (preferably in Python). Besides, you are expected to research and explore innovative ideas. Therefore, you should be able to read scientific papers and adopt code from open-source projects implementing such techniques.

Looking at my situation a couple of years back, I don’t think I would have made a great candidate.

There is of course a lot of room for learning and improvement along the way, but you should be well equipped so you don’t feel overwhelmed throughout the journey. If you don’t feel ready yet, there is nothing wrong about spending a few more months to harness few extra concepts and sharpen your coding skills before you apply. This way you can make the best of the program instead of being discouraged by trying to catch up with your teammates.

Attempting a challenge and submitting your application

Once you are ready for submitting your application, you will need to pick and submit a challenge. There will be few challenges to select from spanning across various tasks like computer vision, NLP, etc. My advice is to pick something you are most familiar and passionate about. This is your time to shine and show your potential.

There are a set of hints for hacking the challenge which are listed by the organizers. Make sure to critically consider those guidelines while developing your solution. The challenge statement might look somehow concise or straightforward (at least it sounded like that to me), but in the end I realized this might have been intentional to leave the floor for innovations rather than spelling out all the required steps. In the next section, I am going to share another important resource that you should take to your advantage.

In addition, you will need to prepare a resume and submit a short video that introduces yourself, your motivation for joining the program, and an overview of your proposed solution to the challenge. This video is supposed to replace an interview. So, take that into consideration while conveying the required points. You don’t need to reiterate the solution to the challenge, rather you can highlight some interesting facts about your approach to the problem.

The program is offered three times a year; make sure to choose the best period that suits your circumstances. The summer session might be a good choice if you are still studying.

The admission happens on rolling bases. Submit as early as possible to maximize your chance of acceptance.

Handy resources

  • AMA sessions: Fellowship.ai regularly hosts online sessions to answer questions related to the program.

    This was a very crucial resource for me. I was almost ready to submit my solution to the challenge when I decided to join one of those sessions. Fortunately, the feedback I got from the session led me to further improve my approach by exploring new techniques that I have not used before. It took me almost three weeks to revise the submission, but this had paid off and I received the offer to join the program in about two weeks time.

  • Guidelines: There is a list of guidelines for submitting a compelling solution on the program’s website. As you progress with your development, revisit those points and you will probably find new ways to enhance your methodology.
  • Fast.ai courses: In my opinion, fast.ai courses are a great way to get started with deep learning. Those courses get you up to speed with foundational concepts of machine learning as well as state-of-the-art techniques in applying deep learning to various tasks in computer vision, NLP, tabular data, etc.

Final remarks

I feel very content with my journey in the program so far. There are a lot of things that I have learned along the way. If you would like to know more how a typical day is for fellows, please stay tuned. I will be sharing my review of the program and highlighting what I have learned once the program ends.