KDA-AI 2026

1st International Workshop on Knowledge-Driven Architectures for AI Systems
Co-located with the 23rd IEEE International Conference on Software Architecture (ICSA) 2026
Amsterdam, Netherlands

Register

Call for Papers

Software architectures are being reshaped by the rapid rise of AI-intensive systems, large language models, and increasing automation in software engineering activities. At the same time, architects face growing challenges related to system complexity, quality trade-offs, ongoing evolution, and sustainability. At the heart of these challenges is architectural and software engineering knowledge, including design decisions, patterns, rules, metrics, models, and assumptions. How this knowledge is captured, represented, reasoned about, shared, and evolved is becoming increasingly important, especially as AI techniques are used both within software systems and to support software engineering itself. This workshop offers a broad and inclusive forum to explore knowledge-driven software architectures in modern systems, with a particular focus on how knowledge-driven architectures and software engineering principles can inform, support, and automate architectural decision-making throughout the software and AI lifecycles. Instead of concentrating on a single paradigm or technology, the workshop welcomes diverse perspectives, methods, and maturity levels, aiming to foster discussion, share early results, and build a vibrant community around knowledge-driven approaches to software architecture in the age of AI.

Topics

Topics include, but are not limited to:

    Architectural Knowledge and Decision-Making
  • Architecture decision records (ADRs) and decision rationale.
  • Architectural patterns, antipatterns, smells, and metrics.
  • Knowledge representation, traceability, and consistency in architectures.
  • Architecture theory, principles, and foundational models.
  • Knowledge-Driven Software Engineering
  • Knowledge-centric approaches to software and system architecture.
  • Rule-based, constraint-based, or model-driven architecting.
  • Architectural knowledge graphs, ontologies, and repositories.
  • Continuous and automated management of architectural knowledge.
  • AI and Architecture
  • Architectures of AI-enabled, data-intensive, and ML-based systems.
  • Use of AI, LLMs, or agents to support architecting activities.
  • AI-assisted design, analysis, validation, refactoring, and documentation.
  • Multi-agent and autonomous architectural systems.
    Architecture Lifecycle and Evolution
  • Architectural refactoring, migration, and modernization.
  • Monolith-to-microservices and service decomposition.
  • Continuous architecting and architecture-as-code.
  • Architectural decay prevention and recovery.
  • Quality Attributes and Trade-offs
  • Performance, scalability, resilience, and reliability.
  • Security, privacy, and trust at the architectural level.
  • Sustainability, energy efficiency, and environmental impact.
  • Cost, operability, and maintainability considerations.
  • Self-Adaptive and Dynamic Architectures
  • Feedback loops and control architectures.
  • Self-managing, self-adaptive, and self-evolving systems.
  • Edge-cloud, distributed, and heterogeneous environments.

Important Dates

Submissions Deadline

February 13, 2026 February 20, 2026

Acceptance Notification

March 13, 2026

Camera Ready

March 27, 2026

Workshop

June 22, 2026 (TBC)


Note: Deadlines are AoE (UTC-12h).

Submissions

All submissions must adhere to the IEEE Computer Science proceedings formatting guidelines and be submitted through the official ICSA submission system. Contributions should be submitted before the deadline using the online platform: EasyChair and must be original, not previously published, and not under review elsewhere.

The review process will follow a single-blind model: reviewers will have access to the authors' identities, whereas the reviewers' identities will remain anonymous to the authors.

Types of Contributions

We welcome high-quality, original contributions in the following formats:

  • Research Papers (8 pages, IEEE format): Novel frameworks, methodologies, or empirical studies.
  • Short papers (2-4 pages, IEEE format): Position papers, preliminary results, tool demos, or experience reports.

Proceedings

Workshop proceedings will be published in the ICSA 2026 Companion proceedings and available in the IEEE Xplore Digital Library. At least one author of each accepted paper must register for the conference and attend the workshop.

Committee

Organizing Committee

Program Committee

Accepted Papers

Title Authors
CellFlow: A Tool For Automatic Jupyter Notebook Workflow Visualization Selin Coban, David Kierdorf, Cagatay Akpinar and Horst Lichter
An Efficient Approach for Model Recovery from Image Containing Diagrams Gopal Shendge, Lisha Ahuja, Chandan Prakash, Pavan Kumar Chittimalli and Ravindra Naik
CAKE: Cloud Architecture Knowledge Evaluation of Large Language Models Tim Lukas Adam, Phongsakon Mark Konrad, Riccardo Terrenzi, Florian Girardo Lukas, Rahime Yilmaz, Krzysztof Sierszecki and Serkan Ayvaz
An Empirical Analysis of LLM-Driven Refactoring for Microservices Abdullah Huzeyfe Köse, Rahime Yılmaz and Feza Buzluca

Program

WIP: TBC...

The KDA-AI 2026 workshop has been joined with the Behaviors 2026 workshop on Architecting Behavior-Aware Software Systems (Website: Behaviors) to make a full day plan.
The program has been composed of presentations of accepted papers, followed by open discussions on the presented work and a shared keynote.

Time Event Speaker(s)/Author(s) Session Chair(s)
09.00-09.10 Welcome and Opening Remarks - Behaviors: Architecting Behavior-Aware Software Systems Behaviors Committee TBD: Mina, Sherkan, or Mahyar
09.15-09.35 Human-in-the-Loop in MLOps: A Systematic Review on Balancing Automation and Human Oversight Louisa Klüter, Dóra Kementzey, Faezeh Amou Najafabadi, Keerthiga Rajenthiram and Ilias Gerostathopoulos TBD: Mina, Sherkan, or Mahyar
09.40-10.00 Toward Behavior-Based Adaptive Decision-Making Systems for Cognitive Cities Rawan Alraddadi, Mohammad Alshayeb, Sajjad Mahmood and Mahmood Niazi TBD: Mina, Sherkan, or Mahyar
10.05-10.25 Human Behaviour as a Security Boundary: Affective AI and Behavioural Analytics for Cybersecurity Zain Akbar, Anilson Monteiro and Jhonny A. Mosquera Ocampo TBD: Mina, Sherkan, or Mahyar
10.30-11.00 Session Break
10.05-11.25 Rewriting the Maze: Developer Experiences Moving from Bare Metal C to Lingua Franca in Embedded CPS Maciej Jakub Grasela, Rahime Yilmaz and Krzysztof Sierszecki TBD: Mina, Sherkan, or Mahyar
11.30-12.25 Open Discussion on Presented Papers TBD: Mina, Sherkan, or Mahyar
12.25-12.30 Closing Remarks and Summary Behaviors Committee TBD: Mina, Sherkan, or Mahyar
12.30-13.30 Session Lunch
13.30-14.00 TBD: Shared Keynote TBD TBD: Mina, Sherkan, or Mahyar and Anders, Mayhar, or Jakob
14.00-14.10 Welcome and Opening Remarks - Knowledge-Driven Architectures for AI Systems KDA Committee TBD: Anders, Mayhar, or Jakob
14.10-14.30 CellFlow: A Tool For Automatic Jupyter Notebook Workflow Visualization Selin Coban, David Kierdorf, Cagatay Akpinar and Horst Lichter TBD: Anders, Mayhar, or Jakob
14.35-14.55 An Efficient Approach for Model Recovery from Image Containing Diagrams Gopal Shendge, Lisha Ahuja, Chandan Prakash, Pavan Kumar Chittimalli and Ravindra Naik TBD: Anders, Mayhar, or Jakob
15.00-15.30 Session Break
15.35-15.55 CAKE: Cloud Architecture Knowledge Evaluation of Large Language Models Tim Lukas Adam, Phongsakon Mark Konrad, Riccardo Terrenzi, Florian Girardo Lukas, Rahime Yilmaz, Krzysztof Sierszecki and Serkan Ayvaz TBD: Anders, Mayhar, or Jakob
16.00-16.20 An Empirical Analysis of LLM-Driven Refactoring for Microservices Abdullah Huzeyfe Köse, Rahime Yilmaz and Feza Buzluca TBD: Anders, Mayhar, or Jakob
16.25-16.55 Open Discussion on Presented Papers TBD: Anders, Mayhar, or Jakob
16.55-17.00 Closing Remarks and Summary KDA Committee TBD: Anders, Mayhar, or Jakob

Contact

Location:

Vrije Universiteit Amsterdam
Netherlands

Call:

+45 6550 4579