01
Curiosity Before Conclusions
Better answers begin with better questions.
Data Science · Machine Learning
SuyogKarki
I build end-to-end ML pipelines — turning messy, imbalanced, real-world data into models. Currently hunting for a data science / ML internship.
Click the portrait to scatter & rebuild
Scroll
About Me
I got into data science by being wrong a lot— my first models looked perfect until I met data leakage. Now I'm obsessed with the questions most people skip: where did this data come from, and is this metric lying to me?
I turn messy, imbalanced, real-world datasets into things people can actually use — and I enjoy the unglamorous 80% (cleaning, validating, doubting) as much as the modelling.
BSc IT — Data Science · Kings College, Kathmandu
Affiliated with Westcliff University, California · 2023 – 2027 · 4th year+2 Science · Kathmandu Model College
2020 – 2022Certifications — click to view
Tech Stack
Python0%SQL0%
Pandas0%
NumPy0%
Matplotlib0%
Seaborn0%
Scikit-learn0%Power BI0%
Excel0%
Git & GitHub0%
Streamlit0%
OpenCV · YOLO · OCR0%Currently learning: Object Detection — OpenCV · YOLO · OCR
Projects on the Ball
Every hexagon on the ball holds one project — click a patch (or a card) to open it.
10 Values I Follow in Data Science & ML
02
Context Before Calculation
Numbers do not speak clearly without context.
03
Quality Before Quantity
More data cannot repair poor data.
04
Simplicity With Purpose
Complexity should earn its place.
05
Evidence Over Assumption
The goal is not to be right — it is to discover what is true.
06
Honesty Over Impressive Metrics
One impressive number can hide many failures.
07
Adapt Without Ego
Keep the objective. Change the route.
08
Explain What I Build
Understanding should travel with the prediction.
09
Remember the Human Behind the Data
Data points may carry real consequences.
10
Stay Unfinished
There is no permanent finish line in a changing world.