commit 648a667 (2022-02-02 13:10:22 -0500) Torsten Scholak: be more precise

Home Page

Hi!

Welcome, my name is Torsten Scholak. I work at ServiceNow Research, where I am a senior researcher and project leader in the Human-Machine Interaction Through Language research program.

My current research interest is deep learning for code, especially with large language models and integrating symbolic reasoning. This year, I am co-organizing the Deep Learning for Code workshop at ICLR. Previously, I have been more focused on semantic parsing. I have published several papers on translating natural language questions into SQL queries. To date, my most successful work on this topic is the PICARD project, which is a state-of-the-art text-to-SQL parser. All my publications can be found here.

I am a main contributor to Hasktorch, a port of PyTorch to the Haskell programming language. I mentored two students in the Google Summers of Code of 2020 and 2021 who worked on the project. I developed an experimental gradually typed API for Hasktorch that optionally tracks the shapes of tensors and operations in the type system. I use Haskell for research, development, and recreation. This website is built with Haskell, Nix, and GitHub Actions.

My path started in Germany, where I grew up and studied theoretical physics at the universities of Bayreuth and Freiburg. My studies concluded in the summer of 2011 when I earned my Ph.D. I then moved to Canada to become a researcher at the University of Toronto. I eventually left academia and started as a software developer and data scientist at a local startup. In 2017, I moved to Montreal and joined Element AI, where I developed machine learning models and solutions for businesses. Element AI was acquired in 2021, and I am now a full-time employee at ServiceNow. My resume is available here.

Bye!

You have reached the end of this page. Before you go, make sure to have visited the links at the top. Also, feel free to reach out to me. I am always happy to hear from people.