Hi! I’m Ivan Zhou

I am a research engineer at Databricks Mosaic. My primary focus is building high performing AI systems to solve real-world challenges and create values.

Previously, I have:

Most Recent Work

  • 2024-Present: Research Engineer, Databricks Mosaic

  • 2023-2024: Machine Learning Engineer, Uber

  • 2018-2023: Machine Learning Engineer, Landing AI

Education

  • MS Computer Science, Stanford University

  • BS Engineering, University of Toronto

More

Featured
Joining Databricks Mosaic
Personal Growth
Joining Databricks Mosaic
Personal Growth

I am thrilled to be joining the Databricks Mosaic team to work on applied AI research. This is a fantastic opportunity for me to continue pushing the boundaries of AI advancement and contribute to building a successful AI platform for enterprises.

Personal Growth
From Vienna to Paris: A Collection of My Favorite Travel Photos in Europe
Journal
From Vienna to Paris: A Collection of My Favorite Travel Photos in Europe
Journal

My girlfriend and I recently traveled to Europe for a conference as well as a vacation. We had a great time walking and exploring each of the cities that we visited. In this post, I am sharing my favorite photos captured in each city.

Journal
My Ray Summit Talk - Building Multi-Modal Foundation Models for Document Automation
Machine Learning
My Ray Summit Talk - Building Multi-Modal Foundation Models for Document Automation
Machine Learning

I gave a presentation at the Ray Summit on my work building Multimodal Foundation Models for Document Automation at Uber. It is always a great pleasure to publicly share what I have been building over the past year!

Machine Learning
Featured
SimpleQA - Build High Quality Benchmark Dataset with LLM + Human

The OpenAI team built a new benchmark dataset called SimpleQA that evaluates large language models' (LLMs) ability to answer factual questions. A particularly intriguing aspect of this paper is, in this era of LLMs, how the team of researchers leverages LLMs in their own workflow to design, iterate, and analyze a new dataset.

Read More →
Thought Preference Optimization (TPO)

Thought Preference Optimization (TPO): Prompt the model to generate a thought process followed by the response. TPO demonstrates significant performance gains on non-reasoning categories, including translation, marketing, and health; reasoning categories like math and analysis also show improvements.

Read More →

Contact

You can contact me at mail@ivanzhou.me