Automotive AI
ADAC: semantic search and RAG across two million documents
I worked on a retrieval system spanning vehicle tests, legal travel guidance, tyre databases, and editorial content, improving benchmark precision from 61% to 84%.
The system I joined
ADAC's corpus combined structured vehicle tests, dense legal guidance, comparison tables, and short editorial articles. A uniform retrieval strategy treated fundamentally different information as if it had the same shape.
The product needed trustworthy conversational answers with source attribution, strong German automotive terminology, and an audio path suitable for hands-free use.
Where I contributed
I worked on corpus classification, chunking and retrieval strategy, reranking, answer guardrails, and the text-to-speech delivery path.
The initial benchmark made the architectural mistake visible: two million indexed documents were not useful when the relevant passage structure had been destroyed during chunking.
Decision trail
Constraints became architecture.
Each decision below connects a concrete limitation to the engineering response and its practical effect.
- 01
Constraint
Uniform 512-token chunks split tables, legal clauses, and vehicle-test sections at arbitrary boundaries.
Decision
Classify documents by content type and apply table-aware, sentence-window, or section-aware chunking to each class.
Impact
Retrieval precision increased from 61% to 84% on the benchmark set.
- 02
Constraint
Approximately 8% of documents had missing or incorrect content-type metadata.
Decision
Combine a structural rules pass with GPT-4o-mini classification only for low-confidence documents.
Impact
The correct chunking strategy was applied to 98.7% of the corpus without using a model on every record.
- 03
Constraint
Dense retrieval alone missed exact model codes, tyre dimensions, and jurisdiction-specific terms.
Decision
Fuse vector and keyword retrieval, then rerank the combined candidates before generation.
Impact
Answers preserved semantic recall while improving exact-match behaviour and source relevance.
Outcome
What the work produced
- Benchmark retrieval precision reached 84%, up from a 61% baseline.
- The pilot resolved 73% of member queries without human escalation.
- Source attribution and guardrails kept answers tied to approved ADAC material.
- Automotive text-to-speech met the target intelligibility threshold with a maintained terminology dictionary.
Technology
RAG quality is often a data-modelling problem disguised as a model problem. The biggest improvement came from respecting the structure of the source material before retrieval began.
Abrar Ahmed