Why I left my very well paid job to become a Data Entrepreneur

Are you confused between Ponicode and CircleCI? It’s not you, it’s us. Ponicode was acquired by CircleCI as of March 2022. The content and material published prior to this date remains under Ponicode’s name. When in doubt: Ponicode = CircleCI.

When you have to make a choice between 2 great options: being a Data Scientist or an Entrepreneur … and you choose both!

Have you ever had to choose between two options you love and abandon one of them? I have, so many times in my life: Engineer or Doctor? Karate or MMA? Sleep or Party? Amelia or Sophia? Burger or Pizza?

But the one time it was impossible to make that choice was when I had to decide between being an Entrepreneur or a Data Scientist.

I first chose Entrepreneurship. The first year after finishing my degree I started 4 businesses one after the other. It was super exciting. I discovered how to give life to an idea and make a business work from scratch. As a fresh graduate engineer who had never done business before it completely changed the way I think and act. It was like adding a new dimension to my brain. I was learning, and making cash on my own. No useless tasks that felt like homework. Everything I did was purely for the benefit of my business. That feeling of freedom and accomplishment was priceless.

However… My daily routine was all about business, marketing, law, accounting. As an engineer I was missing science and technology terribly. And another thing: the products I was selling did not interest me at all and everyday was the same. Slowly, my businesses felt more and more boring and my motivation was dwindling fast.

As a free spirit, I chose to stop my business for a while to get back to what I love: data science. My entrepreneurial experience meant I had a different profile and made it easy for me to find a good job. Within two weeks I was signing my first contract in an amazing digital marketing startup, Hivency as their first data scientist. This was a wonderful experience. The job was stimulating, I was learning a lot, the people were fantastic. I was part of an incredible team, and had a stable and well-paid job with good opportunities for promotion. A promising career was taking shape in front of me.

But deep down inside something was bothering me. It was that need to build something big, something incredible as an entrepreneur. And as long as I had that feeling I could never be fulfilled. Day after day, this feeling became stronger and stronger, gnawing at me despite all my efforts to ignore it.

Then one day my mother introduced me to Patrick Joubert, the entrepreneur who set up the company she works for.

Ponicode for Data Scientists. Ponicode for Entrepreneurs.

Who is Patrick? A successful serial entrepreneur who has the gift of making 4 kinds of miracles come true:

  1. Finding an idea that is fascinating and ahead of the state of the art, especially in data science.
  2. Building a team with very complementary, enthusiastic, resilient, hard-working people. (No, I’m not bragging 😉)
  3. Making everyone who works with him an expert in their fields.
  4. Last but not least, passing on his entrepreneurial mindset and methods.

When Patrick told me about Beamap, and Recast.ai, his two highly successful previous startups, I was amazed. I was even more amazed when he mentioned his new product idea: help developers focus on what they like to do most, on the exciting parts of their job. Creating a unique AI coding partner that will handle all low-interest but essential tasks for them and allow them to focus on what lies ahead. The company would be called Ponicode.

AI on code! It’s AMAZING. I could already see the genius and success that would follow. And applying Machine Learning to generate code is like offering chocolate to a hungry kid. That idea resonated with both my entrepreneurial and data scientist mind, and Patrick sensed that. He asked me to join the founding team of Ponicode.

As a Data Scientist these are the opportunities that I saw in Ponicode:

  1. Make a product that I would personally use day to day.
  2. Work in a broad spectrum of AI fields (Vision for front generation, NLP for code generation, …) and have a chance to become an expert in all of them.
  3. Participate in the whole pipeline (data generation, machine learning, data engineering, data analysis)
  4. Work with a very experienced and successful CEO
  5. Have lots of fun! We are not predicting the number of flowers in a field or the price of houses. We are generating code!!!

As an Entrepreneur, these are the opportunities that I saw in Ponicode:

  1. Learn methods from a highly experienced entrepreneur (Patrick, CEO).
  2. Learn the strategies that can set a company apart from its competitors.
  3. Witness and participate in the evolution of a highly technological product from A to Z.
  4. Be exposed to all the dimensions that make an AI startup successful: Sales, Communication, Legal, R&D, HR, Marketing, Strategy, Product, Finance.

Instantly, I knew that I would be fulfilled both as an Entrepreneur and as a Data Scientist. And I was right. Since day one at Ponicode, that feeling of restlessness has completely disappeared. I have never felt more grounded. I am learning a million things and my mind sharpens everyday. Every week flies by in an instant, but at the same time feels like a year in terms of what I learn and achieve.

I would like to share a few lessons I have learnt in the last 3 months on what it is to be a Data Entrepreneur.

Do Business not Useless

An issue that I have encountered when building my models is getting stuck on irrelevant tasks. Sometimes, your scientific mind and your business mind will contradict one another. You business voice tells you “Do it fast, clients are waiting..” while your scientific mind tells you “This is not perfect! I want 90% of accuracy not 60%!”.

At the beginning, you need to focus on making an MVP as fast as possible. It is not an accuracy contest. You should aim at an MVM, as in Minimal Viable Model. This means taking shortcuts even if it means not picking the state of the art or not listening to senior Data Scientists.

How to know what shortcuts to take?

Always keep in mind that your model is used for your product. If you understand who your product is for, you’ll know exactly where you can make compromises with your data and where you can not.

Also, if you are lucky enough to be surrounded by Tech Entrepreneurs like I am at Ponicode, you’ll be amazed to see how some clever one-hour shortcuts they suggest can save you weeks, if not months.

Don’t underestimate the value of experience

As a young and determined Data Scientist I always thought I could hack my way into any technical challenge. I always underestimated the influence that someone else’s expertise could have on me, and didn’t really feel the need to be guided in my work.

Working alongside my CEO, Patrick, made me realise how wrong I was. Having a constant dialogue with someone of his experience, and being able to exchange views especially when I reach a deadlock has really changed my perspective. The knowledge he is able to share clearly goes beyond mere technical vision and the capacity to manage a team. The most interesting thing is his perspective on how to address complex problems keeping both business and technical needs in mind.

Surrounding yourself with such experience will make you learn and grow ten times faster.

Choose exciting projects

Choosing projects with a product that fascinates you is essential to your fulfilment. Work will stop being a pain and become a passion instead.

You will be happy in the present and your passion will make you learn and perform ten times if not a hundred times better. It is essential to your present and future.

Don’t teach your brain to overfit

One thing that amazed me when I joined Ponicode is how efficient I became by having a very broad view over the project. The more I work in new fields the better I become at learning new things and I even improve at those things I thought I had already mastered.

Also, working on the whole pipeline (data generation, processing, machine learning, data engineering, data analysis, model in production) is highly rewarding, and it is a real advantage if you're aiming towards a successful career in data science.

The bonus for me at Ponicode is that I can inject my creativity into the data. In most companies, very structured processes risk killing your creativity, which results in less efficiency. Creativity makes you discover shortcuts that aren’t taught in those processes. Never forget that your neurons fire the most when you are having fun!

The emotions of an entrepreneur

What I am able to do at Ponicode is make high impact decisions. Each of the decisions I make will have a high impact on the AI and consequently on the product. You have to own up to yourself and your responsibilities. If you succeed, the product lives. If not, it dies. Eat or be eaten.

You enter survival mode. This unleashes your brain and allows you to think of the fastest and most efficient ways to pick the right model, the right data, the right training.

I love survival mode.

Best news is, we are hiring! If you are interested in the perfect balance between entrepreneurship and tech and want to see your learning curve rise through the roof, Ponicode is the perfect opportunity for you.

Join me in this adventure!

Edmond Aouad, Data Scientist at Ponicode

Green blobred blob