From a biomedical lab to AI architect

People often ask me, "Didn't you study biology? How did you end up as an engineer?"

The answer doesn't fit in one sentence, so I started a "Career journey" category to write it out over time. This is the first post, and it lays out the whole route. For the full timeline, see my career path.

1. The start: learning on my own, on a path others expected

I've always liked figuring things out myself. Growing up I entered art competitions and was in a choir and an art class, and I developed an instinctive, stubborn sense for aesthetics and design.

At university I studied food science at National Taiwan Ocean University, which is mostly biology and chemistry. My grades were fine, but the further I went, the clearer it was that this wasn't what I wanted to do long term. I had already started teaching myself web development, though I hadn't thought of it as a career yet.

2. The turning point: my first real system, built in a biomedical grad program

For my master's I joined the Institute of Molecular Medicine and Bioengineering at National Chiao Tung University, working in the dean's computer science lab on gene sequencing and bioinformatics, analyzing data in R and Python every day. My thesis was a testing tool for chronic kidney disease.

The real turning point came in a computer science elective. I built a front-end and back-end integration for a class project, the professor was impressed, and real client work started coming my way.

From there I worked my way from HTML, CSS, JavaScript and PHP to Angular, React, Vue and Express.js. With no computer science degree, I had to build my skills one project at a time.

3. Alternative military service: a second job after hours

During my alternative military service, I spent every evening filling gaps in computer science fundamentals, building a portfolio, and starting my own brand, "1914." One project from that time was a customer service chatbot with intent recognition, connected to LINE, Facebook chatbots and Google Dialogflow and deployed on Heroku.

Looking back, it was the first time I took an idea all the way to something other people could actually use.

4. Into the industry: starting in the back end, across four kinds of work

After my service I deliberately went for a pure back-end engineering role, passed the assessment, and was given plenty of room to explore. Over the eight-plus years since, I've worked roughly in these settings:

Setting	What I learned there
Enterprise systems	Modernizing legacy systems: separate front and back ends, CI/CD, security scanning
E-commerce and SaaS	High traffic and reliability: millions of push notifications an hour, SLAs, blue-green deployment, multi-cloud
Finance and manufacturing	Delivering under strict constraints: container platforms, microservices, integration with large systems such as SAP
International teams	Multi-system integration, and leading the user experience design of a POS

Today I work on AI platform architecture: RAG, on-premises model services, AI agents, and getting these services into real use.

Throughout, I've also taken freelance work across design, front end and back end, which gradually filled out my UI/UX and full-stack skills.

5. Three things crossing fields left me with

If I had to answer "did studying biology help you as an engineer?" in one sentence: yes, and more so the further I go.

1. A lab habit: verify before you believe

In research, a result without a control group that can't be reproduced doesn't count. That habit came straight into engineering: verification steps before launch, and monitoring that tells you what went wrong.

In the age of AI this matters even more. RAG and AI agents produce convincing output, so whether you can evaluate it and prove it's right decides whether it can go live.

2. An art background: interfaces should be understood at a glance

Art training as a kid made me sensitive to whether something looks right and flows well. That became my standard for experience design: a good interface needs no manual, and it should be accessible and use readable type.

3. Teaching myself: nobody teaches you the next technology

Without a computer science degree, I had to find answers and build things myself from day one. The technology keeps changing, from jQuery to React, from virtual machines to Kubernetes, and now to LLMs. What carries over isn't any one framework. It's the ability to learn something new quickly and get it to the point where it can ship.

6. The biggest lesson of these years

Good architecture creates long-term value for a business, far more than the short-term gain of rushing to launch.

I only really understood this after high-traffic e-commerce and the strict rules of finance. Many problems aren't bugs in the code. They come from not thinking through how the data flows and how to roll back when something breaks.

7. Outside of work

I still have a lot of interests: singing lessons, gymnastics classes, and I'm a certified spin instructor. I also enjoy being a couch potato with a good series. For me, the rhythm of life outside work is what keeps the work sustainable.

What I plan to write next

This series will unfold one stage at a time. Topics I have in mind so far:

• How I planned self-study and a portfolio without a computer science degree
• What switching from biomedicine to engineering actually looked like: résumé, interviews and the first job
• What each industry taught me about engineering: e-commerce, finance, manufacturing
• Moving from engineer to architect: what actually changes in the work

If you're considering a career change across fields, or you're curious about any of these topics, subscribe for updates or book a time to talk.