前言 / Introduction

This year, I did something a little scary but really exciting:
I submitted a talk to PyCon TW!

The topic?
From PyTorch to ONNX: Building an Efficient Image Classification API with a Real-Time Web Demo.
Yeah, it’s a long title — but it pretty much sums up everything I’ve learned since college (Computer Vision).

This post is just a chill recap of how I got here, why I wanted to do it, and how AI tools (ChatGPT & Cursor) helped me along the way.

今年我想來點不一樣的，所以就跑去投稿PyCon TW.
下面大概講一下，投稿心路以及AI是如何幫助我的。

Why I Wanted to Submit (a.k.a. why stress myself lol)

2025 marks my third time attending PyCon TW. So I thought, Maybe this time, I don’t want to just sit and listen — I want to share something too.

After a long day of work, I still enjoy building small side projects in my free time.
Then a thought hit me: “Why not talk about the stuff I’ve been building?”

The idea came from a simple motivation:

  • I had a PyTorch model, and I wanted to make it faster and easier to share — via an API and a simple web UI.
  • Then I discovered ONNX + FastAPI + a bit of frontend magic — and it just clicked.
  • Eventually I thought: This might be worth sharing on a bigger stage. And PyCon felt like the perfect place.

想說都參加第三年了(當聽眾也是會膩的)，那要不上台試試。
就利用下班的空餘，寫了一個簡單的side project，把大學到研究所的小知識拼裝起來，拿出玩玩，也順便寫寫一些跟工作無關的，讓大腦放鬆。

How AI Helped Me Get There

To be honest, I had zero frontend experience before this. But I really wanted a small, working demo — so yeah, I needed help. That’s where AI came in. And by AI, I mean my two MVPs: ChatGPT and Cursor.

Here’s how they helped:

  • ChatGPT

    • Cleaned up my messy draft ideas
    • Suggested better writing flow
    • Helped structure the slides

  • Cursor

    • Gave code suggestions
    • Helped me learn new tech on the fly

It honestly felt like pair programming with a tireless teammate who never gets annoyed.

常常被問:身為軟體工程師，怎麼使用AI?

那這邊就是我平時使用的情況:
想法方面，我經常會與ChatGPT討論，因為他總是能夠給與及時的回饋，延伸更多想法，討論是前進中很重要的一環，而他做到了。
而編程方面，則是Cursor和Github Copilot，他倆都不錯，寫程式都是一流，甚至比我還能寫，既乾淨、可讀性又高。人的大腦專心思考邏輯與框架，而AI專心簡化建構的過程，大大加速開發的流程。

並且這也是我第一次寫前端，老實說，AI真的是跨時代的發明，讓許多事情能購輕易地實現。

Thoughts & Feels

I don’t know if I’ll nail the talk or completely blank out on stage.
But honestly, I’m just glad I went for it.

Sharing doesn’t just help others — it helps you understand your journey better.
And who knows? Maybe this blog post will inspire future-me to submit again.

我也不知道會不會投上，反正就當買樂透。
分享一直是我很喜歡做的事情，不僅能讓努力被看見，也能讓缺點被發現。

Curious about the actual project?
You can check out the code and demo setup here:

Feel free to explore, try it out, or leave a star ⭐ if you found it helpful!