Python’s influence is growing with AI and I am always very happy about it. Python’s been my favourite programming language due to its simplicity. My own Python journey began with web backend development using Django, which is a vast domain in itself. It took me a while to get to Python’s usage in data science.
As I see it, Python’s usage broadly comes down to two major applications in the industry:
- Web backends and
- data processing, AI, and machine learning
I started enjoying building with Django instantly, starting with my first project in it. I have used it for numerous projects by now. Django helped me get POCs out very quickly (And no, its performance never truly became a bottleneck). Yet, for the longest time, the sheer scale and complexity of Python on the data science, Data Engineering, and AI/ML side remained almost a black-box. It wasn’t until I genuinely dove in that I now hopefully have a comprehensive understanding of the language.
Interestingly, I’ve often observed more robustly structured Python code in the web backend world. Perhaps this stems from the structured nature of web frameworks, the prevalence of MVC patterns, and a more pronounced use of OOP and classes. Backend systems also impose strict requirements for maintenance, uptime, and scalability, which I believe inherently drives more rigorous project development. I genuinely don’t see one side trumping the other. My love for a well-crafted backend system built with Django/FastAPI is as strong as my appreciation for Python’s power in data analysis and AI.
Backend technologies, viz. webservers, containerization (Docker), Kubernetes, Nginx, and cloud (AWS) are great to keep in your toolbox, whichever side you are on: it form the fundamental blocks in operationalizing ML models in production.
Leave a Reply