Get "Alertyd" — My First Project Soon to Come

Picked it up after a month

Because I insisted, Claude allowed me to work on my own project before I devote myself into Speech Processing. I wouldn’t say this is procrastination but responsibility. Because one should finish what he promises his friends — a product that delivers the hottest summarised (important and fast) news of users’ choice to their email. Whether you want to follow the movement of your business opponent or of your idol, Alertyd will keep you away from falling behind.

Stack

Considering the type of service, one which requires automated emails and content crawling, I decide that the frontend and backend of this product can be separated. From what I know, it is more common for an indie developer to do everything in one sweep using frameworks such as Next.js, go, or something else that I don’t know yet. Well, I guess there’s no need for it if I don’t even demonstrate this product on the webpage.

For frontend, which really isn’t that much of a trouble these days with the help of IDEs like cursor, I decide to go with Next.js, but I haven’t started yet.

I don’t really know programming, but if I have to say that I know one programming language or else I’ll be murdered, I’d say I know Python. For the same reason, I choose Python as my backend language. I made this decision for two reasons:

1. The languages aren’t that different, so choose the most familiar one. Their are no fundamental barriers between languages especially at this level, meaning that any language will do the job. Also, I’ve heard that it’s better for a beginner to master one coding language than knowing ten but not being good at them. One would certainly consider he who can do anything with python a better programmer than he who can write “Hello World!” in 20 languages. So, I should use the language that I’m familiar with to save me the some trouble.
2. Python is popular. Though it might not be the most suitable for every job on earth (certainly not game developing), but as far as I’m concerned, most of the fields that I’m interested (AI, Computational Linguistics…) are fairly compatible with Python. I assume writing Python even with the help of AI, will help me learn it. A better grasp of Python will help me with my future career in most of the areas. This increases the reuse rate of my time as I can both do what I like and learn Python.

To create a deployable project I also need to choose a framework. When I was in the age of naïveté and innocence, I used to write an entire application on Jupyter Notebook. From what I know, you can’t deploy a Jupyter Notebook very easily. So I did some research and found that there are two options for me, Flask and FastAPI.

I chose Flask at first because it’s supposed to be easy and with better support from the society. But the way I used it doesn’t seem right to me. I made it too messy that I wanted to start again.

I finally go with FastAPI, because it’s supposed to be faster and is more dedicated for API?

Now it’s mostly functioning, and certainly looks better after cleaning.

What to do next?

There is one thing. I use jina for web crawling, but I’ve run out of free token and can’t afford it. So now I decide to build frontend and think of another way afterwards.