Make It Yours: A Better Way to Build Your AI Portfolio
Why solving your own problems leads to the most impactful projects.
“Which portfolio project should I build for my AI career?”
I hear this question all the time—especially from my mentees.
Yes, there are plenty of GitHub links and curated lists of AI project ideas floating around on LinkedIn. They’re good for learning. But they’re not enough if your goal is to stand out.
Here’s the problem: If you're just cloning someone else's project, you’re only proving that you can follow instructions. That’s a solid start—but not a signal of creativity or problem-solving.
Recently, I spoke with a mentee who wanted to build an AI agent. I asked her to take an AI Agent course from DeepLearning.ai. As she worked through the exercises, I gave her this challenge:
“Out of all the ideas that come to mind after taking the course, which one could solve a personal problem for you?”
That changed everything.
She’ll use the same tools and techniques—but apply them to her own use case, rooted in her job search. That project will become uniquely hers. It won’t be just a demo. It’ll be useful. And personal.
A Personal Example: My Audio-Ready Research Article Reader
Here’s an example from my own experience:
Ever tried listening to a research PDF while walking, only to waste minutes skipping over dense experiment sections?
I have. Many times.
I use read-later apps like Matter to turn PDFs into audio. But not all sections are worth hearing out loud—especially when walking or driving.
So I thinking of building a small AI agent that:
Takes in a PDF I plan to read later
Parses its layout using tools like Docling
Removes sections I usually skip (experiments, references, etc.)
Summarizes the intro and methods
Sends the trimmed-down version to Matter, ready for audio mode
It’s not about full summarization. It’s about keeping what I actually want to listen to. A small, personal automation that could save me hours.
How to Choose Your Next AI Portfolio Project
So if you’re preparing for an entry-level AI role, here’s my advice:
➊ → Start with any project that sparks your interest.
➋ → Replicate it. Learn how it works.
➌ → Then adapt it to solve your own problem. Something that would help you.
➍ → You become the first user. That’s when it clicks.
It’s no longer “just another portfolio project.” Now it’s your story.
Got an idea for a personal AI project? Feel free to run it by me—glad to help.
#AIcareers #PortfolioProjects #LearnByBuilding