About me

Hi, I'm Johannes Härtel. I'm a postdoctoral researcher in computer science at Vrije Universiteit Amsterdam. I work on AI-driven software development.

The field moves fast. Not long ago, predicting the next API call was considered progress. Today, agents write code and orchestrate entire software engineering stacks. I'm interested in what this means for how we build and evaluate software.

My current focus is on three things:

• Empirical practice. AI research in software engineering rests on statistics that are often used out of habit. Instead of asking "is this test correct?", I'm more interested in testing how our methods behave under controlled assumptions.
• Simulation as evaluation. If the evaluation is wrong, everything built on top of it is noise. Simulations are transparent, controllable, and repeatable. If we don't fully understand the AI, we should at least understand the world we test it in.
• Agentic coding. How to use coding agents, how to build them, and what changes when prompts become our models. Sometimes this means using agents to write other agents. It's surprisingly effective.

Before joining Vrije Universiteit Amsterdam, I worked on testing cloud-native systems at Vrije Universiteit Brussels. I received my PhD in computer science in Koblenz, where I focused on mining and understanding software engineering data.

You'll find selected past work and teaching below.

Teaching

• Software Threat Analysis (Vrije Universiteit Amsterdam, 2025): I am running the Software Threat Analysis Course at Vrije Universiteit Amsterdam.
• Capita Selecta in Software Engineering (Vrije Universiteit Brussels, 2023/24):  Exercises on bug prediction, running live coding sessions, and simulations. Content available on 2023/24.
• Capita Selecta in Software Engineering (Vrije Universiteit Brussels, 2022/23): Content for the previous version is available on 2022/23.
• Big Data (University Koblenz-Landau, 2022): Big Data course with weekly live coding sessions on Scala, Spark, and AWS DynamoDB.
• Data Science (University Koblenz-Landau, 2021-22): Online course on modern data science.
• MSR (University Koblenz-Landau, 2020): Flipped classroom course on Mining Software Repositories.

Publications

• Yunior Pacheco Correa and Coen De Roover, Johannes Härtel: The sampling threat when mining generalizable inter-library usage patterns Sci. Comput. Program. (2026)


• Johannes Härtel: Improved Labeling of Security Defects in Code Review by Active Learning with LLMs EASE (2025) PDF


• Serge Demeyer and Coen De Roover and Mutlu Beyazit and Johannes Härtel: Threats to Instrument Validity Within "in Silico" Research: Software Engineering to the Rescue ISoLA (2024)


• Ward Muylaert and Johannes Härtel and Coen De Roover: Symbolic Execution to Detect Semantic Merge Conflicts SCAM (2023)


• Johannes Härtel and Ralf Lämmel: Operationalizing Validity of Empirical Software Engineering Studies Empirical Software Engineering (vol 28) (2023) PDF


• Johannes Härtel and Ralf Lämmel: Operationalizing Threats to MSR Studies by Simulation-Based Testing MSR (2022) PDF


• Marcel Heinz and Johannes Härtel and Ralf Lämmel: Reproducible Construction of Interconnected Technology Models for EMF Code Generation Journal of Object Technology (vol 19) (2020)


• Juri Di Rocco and Davide Di Ruscio and Johannes Härtel and Ludovico Iovino and Ralf Lämmel and Alfonso Pierantonio: Understanding MDE projects: megamodels to the rescue for architecture recovery Software and Systems Modeling (vol 19) (2020)


• Johannes Härtel and Ralf Lämmel: Incremental Map-Reduce on Repository History SANER (2020)


• Philipp Seifer and Johannes Härtel and Martin Leinberger and Ralf Lämmel and Steffen Staab: Empirical study on the usage of graph query languages in open source Java projects SLE (2019)


• Johannes Härtel and Marcel Heinz and Ralf Lämmel: EMF Patterns of Usage on GitHub ECMFA (2018) PDF


• Juri Di Rocco and Davide Di Ruscio and Johannes Härtel and Ludovico Iovino and Ralf Lämmel and Alfonso Pierantonio: Systematic Recovery of MDE Technology Usage ICMT (2018)


• Johannes Härtel and Hakan Aksu and Ralf Lämmel: Classification of APIs by hierarchical clustering ICPC (2018) PDF


• Johannes Härtel and Lukas Härtel and Ralf Lämmel and Andrei Varanovich and Marcel Heinz: Interconnected Linguistic Architecture Art Sci. Eng. Program. (vol 1) (2017) PDF


• Simon Schauss and Ralf Lämmel and Johannes Härtel and Marcel Heinz and Kevin Klein and Lukas Härtel and Thorsten Berger: A chrestomathy of DSL implementations SLE (2017)


• Johannes Härtel and Lukas Härtel and Ralf Lämmel: Test-Data Generation for Xtext (Tool Paper) SLE (2014)


• Harbusch, Karin and Cameran, Christel-Joy and Härtel, Johannes: Underspecification-based grammatical feedback generation tailored to the learner's current acquisition level in an e-learning system for German as second language Proc. of EUROCALL (2014)


• Karin Harbusch and Johannes Härtel and Christel-Joy Cameran: COMPASS III: teaching L2 grammar graphically on a tablet computer Proc. of SLaTE (2013) - Workshop speech and language technology in education (2013)


• Harbusch, Karin and Cameran, Christel-Joy and Härtel, Johannes: Personalized grammar teaching on a tablet in inclusion classrooms Proc. of International Conference on Using New Technologies for Inclusive Learning (2013)