Skip to content
Portrait of David Procházka

David Procházka

Ph.D. candidate at the Faculty of Informatics, Masaryk University.

I research how indexes for vector search scale to hundreds of millions of objects, and how they cope when the data keeps changing. With colleagues, we build that research into public search engines, including AlphaFind for protein structures and DreaMS Search for mass spectra.

Publications

  1. 2026
    Recovering Graph ANN Search on Attention-Derived Workloads via a Query-Agnostic Spherical Transformationto appear

    David Procházka, Vlastislav Dohnal, and Martin Aumüller

    Conference on Information and Knowledge Management · Rome, Italy

    Code →

  2. 2026
    Learned k-NN Graph Construction and Spherical Indexing for MIPSto appear

    David Procházka, Emma Sommerová, Jan Mach, and Vlastislav Dohnal

    Similarity Search and Applications · Brno, Czechia

    Code →

  3. 2026
    Capacity-Aware Cluster Rebalancing for Efficient R-Tree Indexingto appear

    David Procházka, Mirjam Bayer, Daniyal Kazempour, Maximilian Verwiebe, Peer Kröger, and Vlastislav Dohnal

    Similarity Search and Applications · Brno, Czechia

  4. 2026
    AlphaFind v2: similarity search in AlphaFold DB and TED domains across structural contexts

    Terézia Slanináková, Adrián Rošinec, Jakub Čillík, Aleš Křenek, Katarína Grešová, Jana Porubská, Eva Maršálková, Jaroslav Olha, David Procházka, Lukáš Hejtmánek, Vlastislav Dohnal, Karel Berka, Radka Svobodová, and Matej Antol

    Nucleic Acids Research

    Website →

  5. 2025
    On the Costs and Benefits of Learned Indexing for Dynamic High-Dimensional Data

    Terézia Slanináková, Jaroslav Olha, David Procházka, Matej Antol, and Vlastislav Dohnal

    Big Data Analytics and Knowledge Discovery · Bangkok, Thailand

    Extended version →

  6. 2024
    Scaling Learned Metric Index to 100M Datasets

    David Procházka, Terézia Slanináková, Jozef Čerňanský, Jaroslav Olha, Matej Antol, and Vlastislav Dohnal

    Similarity Search and Applications · Providence, RI, USA

    Code →

  7. 2024
    AlphaFind: discover structure similarity across the proteome in AlphaFold DB

    David Procházka, Terézia Slanináková, Jaroslav Olha, Adrián Rošinec, Katarína Grešová, Miriama Jánošová, Jakub Čillík, Jana Porubská, Radka Svobodová, Vlastislav Dohnal, and Matej Antol

    Nucleic Acids Research

    Website →Code →Data →Preprint →

  8. 2023
    SISAP 2023 Indexing Challenge – Learned Metric Index

    Terézia Slanináková, David Procházka, Matej Antol, Jaroslav Olha, and Vlastislav Dohnal

    Similarity Search and Applications · A Coruña, Spain

    Code →

  9. 2021
    Organizing Similarity Spaces Using Metric Hulls

    Miriama Jánošová, David Procházka, and Vlastislav Dohnal

    Similarity Search and Applications · Dortmund, Germany (virtual)

    ★ Best Student Paper Award

    Code →

Projects

  1. 2026
  2. 2025
  3. 2024
    Upgrading AlphaFind With High-Quality Embeddings

    Principal investigator · 2024–2025

    Masaryk University

    MUNI/33/1014/2024

Show 2 more ↓ Show less ↑
  1. 2024
  2. 2023

Research stays

  1. 2025
    IT University of Copenhagen

    Algorithms Research Group · hosted by Assoc. Prof. Martin Aumüller

    Copenhagen, Denmark · September – October 2025

  2. 2024

Theses

  1. 2026
    Scalable Learned Indexing for Complex Data

    David Procházka · supervisor Vlastislav Dohnal

    Advanced Master's thesis (RNDr.), Masaryk University

  2. 2023
    Advancing Motion Words for Human Motion Classification

    David Procházka · supervisor Vlastislav Dohnal

    Master's thesis, Masaryk University

    Code →

  3. 2021

Thesis supervision

  1. 2026
  2. 2026
  3. 2026
Show 8 more ↓ Show less ↑
  1. 2026
  2. 2026
  3. 2026
  4. 2026
  5. 2025
  6. 2025
  7. 2025
  8. 2024

Thesis consultation

  1. 2026
  2. 2026
  3. 2026