Technical debt, security vulnerabilities like Log4Shell, and high maintenance costs create significant challenges when updating older applications.
The open source Konveyor project performs static code analysis to create an application inventory and generate detailed reports on migration risks and effort.
A live demonstration shows how Konveyor AI uses a large language model to automatically convert a legacy Java Message Service (JMS) bean to a modern reactive messaging implementation.
The tool automatically transforms a remote Enterprise JavaBean (EJB) that uses the RMI-IIOP protocol into a standard, modern REST API endpoint.
Konveyor AI leverages Retrieval-Augmented Generation (RAG) with static analysis data to provide context to any LLM, using agents to compile and validate the generated code.
The developer workflow involves checking out code, configuring migration targets, running the analysis, and applying the AI-generated patch to complete the migration.
41:17Syed M Shaaf
33:14Daniel Oh
08:23Julia Kordick
15:26Gregor Schumacher, Sujay Joshy & Marcel Gocke
31:02Daniel Oh
32:13Cedric Clyburn
28:09Jan Oberhauser, Mirko Novakovic, Alex Laubscher & Keno Dreßel
29:18Daniel Oh & Kevin Dubois
Jobs that call for the skills explored in this talk.
Amdocs
Kontich, Belgium
Akros Ag
Akros Ag