ABOUT
My name is Kha Vo, an AI scientist / engineer.
My interest of work is to develop creative and deep AI solutions to solve complex problems. That could be any of LLM finetuning, deep learning modeling (time-series, computer vision, NLP, reinforcement learning, graph neural networks…), deep program synthesis, mathematical optimization, AutoML, to statistical analysis. I am also familiar with standard data pipelines, infrastructures, and processes (Azure, Snowflake, SQL, AWS, GCP,…)
My strength is my versatility, depth, and wide exposure to various challenging AI problems, solidly and authentically demonstrated by achieving 9 Kaggle competition gold medals! Out of the best, I finished rank 1/2600 teams in the Google Brain Ventilator Pressure Prediction, and rank solo 4/1400 in the prestigious $1M-prize ARC-AGI 2024. I’m proud to present my Kaggle Achievements and my Kaggle profile.
I would like to briefly explain why “Kaggle” has its own section in my website. In the data science community, a Kaggle competition gold medal is something that is of utmost glory. The reason is that for each featured prized competition, it usually lasted for a few months. So a Kaggle competition is much more competitive than a normal a-few-days hackathon, or a-few-hours competition. To perform well in such lengthy contests, one would need to completely commit all his energy, knowledge, and skills, to continually improve the models, digging into the data, creatively design the solving methods, in order to rise to the top 10 (for the gold medal) of thousands of other competitors.
That also explains why so many top Kagglers are being hired by big techs. For instance, NVIDIA is now hiring roughly 15 out of 30 top worldwide Kagglers.
I am also the top-1 Australian on this platform, with the most competition gold medals that I relentlessly fought for in the last 8 years.
Besides deep data science exposure, I’m also capable of writing softwares or codes in other coding languages, such as Kotlin, JS, and C++. Proud to mention a few projects such as:
- A TSP solver optimized to crack the Kaggle Santa 2018 competition was written in Kotlin (as well as Python). Code here
- A few C++ projects to optimize performance on Kaggle competitions.
- Some of the constellation visualizations (the moving particles on this website) were written in JS. Code here
Occasionally I write Posts in this website, or other short write-ups on Kaggle, to share my technical insights and discoveries.
I have worked as key roles for Collective[i] (data science manager, data science lead), H2O.ai (senior customer data scientist), Oracle (senior applied scientist), CBA (senior data scientist), UNSW (applied science researcher), UTS (post-doc). I received a PhD degree in machine learning and data science from the University of Technology Sydney (UTS).
Also, I have a few notable research Papers.
Besides work, I also enjoy music (check out my Tunes page), chess, soccer, and making funny AI stuffs :).
