quality_assurance · finance · workflow
Adyen augments unit test generation with LLMs, AST parsing, and Knowledge Graphs
Writing unit tests is time-consuming and manual processes introduce variability; conventional LLMs and RAG systems with vector embeddings cannot capture the interconnected, context-dependent nature of large codebases, making automated unit test generation inaccurate.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Unit test need initiated
The need to augment the process of creating unit tests initiates the exploration of LLMs for code generation.
Tools used
LLMsASTKnowledge Graphsgraph databaseRAG
Outcome
Combining AST parsing and Knowledge Graphs with LLMs produced a tangible advancement in unit test generation, streamlining the process and bringing new depth and accuracy, and enhancing developer capabilities at Adyen.
What failed first
Conventional RAG systems using vector embeddings fell short for code generation because they capture only surface-level textual similarities, missing state management, dependency hierarchies, and the context-dependent meaning of code constructs.
Results
Volumeelevating the LLM's ability to generate unit tests accurately
Grounding & classification
Source type: technical build writeup
23 fields verified against source quotes.
code generationragcode diff prknowledge basefailure mode describednamed customerproduction runtime claimedsource backedtools describedworkflow describedfinancial servicesaccuracy improvementemployee productivitytechnical build writeupquality assurancerag answering