Hi, I'm Feng 👋

I'm a data scientist who believes the best work happens when you take what you do seriously, but not yourself too seriously.

By day, I'm diving deep into data and building models that (hopefully) make sense. By weekends, I'm a professional fisher—you'll find me on boats chasing the next big catch, because apparently I enjoy challenges that require both patience and questionable optimism.

My life's been a bit of a global ping-pong game: Guiyang → Guangzhou → Saint John → Townsville → Bimini → Saint Andrews → Toronto → Cambridge → Zaragoza → Nanjing. Each move taught me something new about data patterns, fishing techniques, or how to order coffee in yet another language. Turns out, marine animals behave differently everywhere, but data scientists are universally caffeinated.

Check out my projects and blog below :)

Check out my projects

  • A turtle swimming underwater with a blue background and sandy ocean floor

    Deep Sea Detective

    This AI detective for AR goggles identifies marine life in real-time. Using computer vision, it names species as you dive, turning the ocean into an interactive learning adventure.

    Project Page
  • Flowchart comparing standard, OpenAI LLM automated, and LangChain AI orchestrated business insight pipelines with legend for human judgment and automatic processes.

    Robot Data Analyst

    From business questions to board-ready strategies in under three minutes. I built a robot analyst using LLMs to autonomously perform deep data analysis, delivering actionable insights without human intervention.

    Project Page
  • A circular diagram illustrating ecological impacts on different groups in an ecosystem, with segments labeled as Resident, Fish, Bird, Benthic Invert, Predator, P. Producer, Omnivore, Herbivore, Detritivore, Microbe, Phytoplankton, Zooplankton. Blue segments indicate a decrease in impact, orange segments indicate an increase, and gray segments show no net change. The diagram is surrounded by icons representing various organisms and ecological groups, with a legend explaining impact colors and a label for NNS.

    Global Invasion Decoded

    I tamed chaotic ecological data. By meta-analyzing hundreds of studies, I created a global 'rulebook' for invasive species, helping conservationists predict and prioritize threats before they become a crisis.

    Project Page

Latest Blog

A line graph titled 'Learning Curves: Your Model's Report Card' showing training loss and validation loss decreasing over epochs, used to assess AI training progress.

Learning Curves: Your Model’s Report Card

Is your model learning or just memorizing? This guide treats learning curves as your model’s report card , decoding the patterns that signal success, overfitting, or failure before you waste compute.

A woman smiling and holding a coffee cup while walking on a city street during the daytime, with buildings and trees in the background.

Checkout my

Github Linkedin Google Scholar

If you’ve found your way here, chances are you already have my contact. But in case you arrived by accident—or through some divine algorithmic intervention—hi there!

You can reach me at: fengtang07 at gmail.com

I’m always up for a good conversation about data science, machine learning, marine life, stand-up comedy, or any odd and intriguing ideas.

Let’s connect!