PeopleofSuperintelligence

Researcher

Tim Hempel

Tim Hempel is a research software engineer who specializes in AI for computational molecular biology and coauthored BioEmu, a Microsoft Research AI for Science deep-learning model of protein structure ensembles. In 2021, at Freie Universität Berlin, he was first author of a PNAS paper on independent Markov decomposition and of a paper on the drug candidate otamixaban against SARS-CoV-2 cell entry. A 2022 Nature Communications paper with Andreas Mardt, Cecilia Clementi and Frank Noé joined that decomposition with VAMPnets, a deep learning approach to Markov modeling. BioEmu-1 became open source in February 2025, and the BioEmu paper appeared in Science in July 2025 with Hempel as second author. He is a Senior Research Software Development Engineer at Microsoft AI for Science in Berlin.

Now at
Microsoft
Badges

Timeline

  1. 2021

    Independent Markov decomposition

    In 2021, he is first author of a PNAS paper that introduces independent Markov decomposition, a method that uses weak coupling between subsystems to compute a global kinetic model without sampling every combination of subsystem states. His affiliations are the Department of Mathematics and Computer Science and the Department of Physics at Freie Universität Berlin.

  2. August 2021

    Otamixaban against SARS-CoV-2 entry

    In August 2021, he is first author, with Katarina Elez, Nadine Krüger and others, of a paper showing that the late-stage drug candidate otamixaban inhibits SARS-CoV-2 cell entry and suppresses TMPRSS2 activity in a human lung cell line.

  3. October 2021

    The Deeptime Python library

    In October 2021, he is a coauthor, with Moritz Hoffmann, Martin Scherer, Frank Noé and others, of Deeptime, a Python library for machine learning dynamical models from time series data. The library offers Markov state models, Hidden Markov Models and Koopman models.

  4. November 2022

    Learning independent Markovian domains

    In November 2022, he and Andreas Mardt, Cecilia Clementi and Frank Noé publish the Nature Communications paper "Deep learning to decompose macromolecules into independent Markovian domains". It combines independent Markov decomposition with VAMPnets, a deep learning approach to Markov modeling, and Hempel designs the method and develops the test systems.

  5. February 2025

    BioEmu-1 goes open source

    In February 2025, he coauthors the Microsoft Research blog post that introduces BioEmu-1 as open source, signing as a Senior Research Software Development Engineer. The deep-learning model generates thousands of protein structures per hour on a single graphics processing unit.

  6. July 2025

    BioEmu in Science

    In July 2025, Science publishes "Scalable emulation of protein equilibrium ensembles with generative deep learning", on which Hempel is second author after Sarah Lewis. BioEmu emulates protein equilibrium ensembles by generating thousands of statistically independent structures per hour on a single GPU.

  7. July 2025

    A broad coronavirus inhibitor

    In July 2025, he is second author, after Katarina Elez, of a Nature Communications paper that combines molecular dynamics simulations with active learning to screen drug candidates. The framework cuts computational costs about 29-fold and finds BMS-262084, a TMPRSS2 inhibitor with an IC50 of 1.82 nM.

  8. As of October 2026

    Engineer at AI for Science

    He works as a Senior Research Software Development Engineer at Microsoft AI for Science in Berlin, specializing in AI for computational molecular biology with a strong emphasis on training data.

Career

  1. NowMicrosoftSenior Research Software Development Engineer
  2. Until 2021Freie Universität Berlin

From the biography.

  1. 16% of AI career moves in the 2020s went from a university to a companyIncludes Tim Hempel. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  2. Among employers with 30+ full-time cases, OpenAI leads primary technical role H-1B pay at $310,000, and Amazon, Google and Microsoft file 57% of all casesIncludes Microsoft. Across 440,669 certified US H-1B filings from 44 employers (FY2020–FY2026), OpenAI leads primary technical role base pay at $310,000 and xAI leads overall base pay at $290,000.
  3. Google had the most people shaping AI every year from 2014 to 2025Includes Microsoft. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  4. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Microsoft. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Tim Hempel yet. Send us a free-licensed photo.

  1. Independent Markov decomposition: Toward modeling kinetics of biomolecular complexes pmc.ncbi.nlm.nih.gov. Checked 20 October 2026.
  2. Synergistic inhibition of SARS-CoV-2 cell entry by otamixaban and covalent protease inhibitors: pre-clinical assessment of pharmacological and molecular properties bifold.berlin. Checked 20 October 2026.
  3. Deeptime: a Python library for machine learning dynamical models from time series data arxiv.org. Checked 20 October 2026.
  4. Deep learning to decompose macromolecules into independent Markovian domains nature.com. Checked 20 October 2026.
  5. Exploring the structural changes driving protein function with BioEmu-1 microsoft.com. Checked 20 October 2026.
  6. Scalable emulation of protein equilibrium ensembles with generative deep learning microsoft.com. Checked 20 October 2026.
  7. Simulations and active learning enable efficient identification of an experimentally-validated broad coronavirus inhibitor nature.com. Checked 20 October 2026.
  8. Tim Hempel at Microsoft Research microsoft.com. Checked 20 October 2026.