Focus on quality, not quantity - DSBoost #33
Discover more from DSBoost
Focus on quality, not quantity - DSBoost #33
💬 Interview of the week
This week we interviewed Marving, who is the founder of Dataaxy.
Enjoy:
You have an eclectic background that spans Project Management, Business, and Digital Marketing, along with Computer Science. Could you elaborate on how these diverse skills have shaped your current role, particularly how your expertise in Project Management, Business, and Digital Marketing has influenced your journey?
Those diverse skills influenced it quite a lot! I think a good entrepreneur is more of a generalist than a specialist. To be able to build and make a startup or side project works, you need to be kind of a swiss knife.
Besides, Dataaxy is not reinventing the wheel but addressing an unaddressed niched problem. Which means it’s already a very competitive market. And in a competitive market, you cannot strive without Digital Marketing.
We noticed that your background doesn't include formal training in Data Science or Machine Learning. What motivated you to create Dataaxy, a specialized job board for data professionals?
Along my journey, through IoT, information systems Consulting, I have realized how important big data is and would be. Dataaxy is my way of getting closer to the data field, learning more in Data and ML while building a business around it.
Are you considering diving into Data Science in the future? How do you typically identify and pursue opportunities for professional development?
Yes, I intend doing so. Potentially doing a bootcamp to quickly grasp fully the world of Data Science.
What was the driving force behind the creation of Dataaxy, and how does the platform aim to transform the hiring landscape for data and AI professionals?
Dataaxy was born from recognizing a gap in specialized hiring for Data and AI roles, back in October 2022. Our goal is to streamline recruitment, enabling companies and professionals to connect more effectively as a phase 1. However, we plan on going further with enabling an even more seamless recruitment process for both the Recruiter and the Talent with a niched saas reduction repeating task at every single step of the recruitment customer journey.
How does Dataaxy set itself apart from other job boards, particularly when it comes to accommodating freelance and contract roles?
Firstly, with being niched. No job board niched in Data & AI and stands as a real business and brand and seamless UI exist at the moment. On top of that, we decided to open it as a reverse job board feature uniquely empowers freelancers and contractors, allowing them to showcase their expertise and have businesses approach them directly.
Could you discuss how Dataaxy utilizes data analytics and AI to enhance the user experience on the platform?
As stated above, we aim at improving the recruitment & hiring process at every step of the way. For example, we are working on features that will somehow nudge the talent in that direction such as having a profile helper that will review and enhance the profile of the user based on his research.
What's next for Dataaxy? Are there any new features or updates that you're excited to share with us?
We are currently launching to see how the market react to this state of our product, and based on the feedback we will prioritize the features that we have in our pipelines!
Have you observed any notable differences between the American and European job markets for data professionals? What markets are you currently operating in?
The American market often emphasizes rapid scalability and cutting-edge technologies, while the European market showcases a balance between tech and its societal impact.
Our market is motsly North America being US & Canada.
How do you envision the role of data science changing in the job market, and what strategies does Dataaxy have in place to adapt to these shifts?
It seems like the role of data science is increasingly intersecting with other fields, from marketing to healthcare, to education… With the public revelation of how AI is changing our world, we can only expect the intersection to drastically increase…
You've mentioned that the concept for Dataaxy was born during a workshop at Le Wagon Montreal. Could you share a defining moment or key insight from that workshop that catalyzed the creation of Dataaxy?
Most definitely. I was willing to do something in regard to the Data field, but it’s only when I’ve got a workshop from Joe Masilotti, a Senior Software Engineer in ruby, that it’d clicked. He created a reverse job board for Ruby on Rails Developer, and I said that’s it. That’s how we put our first step in the industry.
🧵 Featured content
How should my projects be to get an internship or job offer?
🗣️ Response from a manager on Reddit:
I like when candidates:
Have original/unique projects. Titanic or mnist classification for example are overused. Come up with your own problem and solve it.
Tried different approaches to solve the problem, can explain their discoveries, wins/loses. This shows you understand the experimental nature of DS.
Model performance isn't very important
Deploy the model somewhere. Newbies often overlook this, but actually productionalizing something will make you stand out. You may even demo it during an interview if manager is interested enough.
Really understand what they've done. Used random forest? Ok, now explain how exactly RFs work, their pros/cons, why you've chosen RF, etc.
Document the project. If I visit your github, I want so understand from the readme what the project is about and what did you do. Not too detailed, just on a high level.
Also, focus on quality, not quantity. It's better to properly understand 5 models and have 3 good projects than 20 models which you can't explain and 10 projects that I've seen already.
Learning the theory is easy, but something that beginners commonly miss is practice with real-world examples. That is how you can really learn Data Science.
As the previous person shared on Reddit, come up with a problem that interests you and solve it! Iterate with multiple approaches, which is often the case in the field and will showcase you are used to dealing with uncertainty.
🗣️ From another Reddit response:
Find something interesting in life. Find data from that thing.
Make the machine learn something interesting from that data. Something that you will find useful or interesting, and may want to share.
You enjoy looking at the weather? Make a weather-learning machine.
You watch sports? Make some predictions using a machine.
Don’t focus on getting the absolute best performance, that’s not the point of these projects, focus instead on really understanding what you are doing and why you are doing it.
Finally don’t forget to deploy it and clearly document your code and project following best practices.
How to get this practical experience if you don’t have time to do a project from scratch? Check the “Do It Yourself” or DIY section of MLPills:
You don’t know how to deploy your model or how to deal with real-world data? Then check Real-World Machine Learning, by :
