TSWAP: A Multilingual Retrieval-Augmented Thai Wellness Advisor
2026-08-24 • Computation and Language
Computation and LanguageInformation Retrieval
AI summaryⓘ
The authors developed TSWAP, a chatbot that gives wellness advice in eight languages, based on a verified knowledge base of Thai traditional medicine and certified providers. They use a language model combined with a specialized retrieval system to find accurate information and apply safety rules to avoid inappropriate medical advice. The system was tested with benchmarks and real user data, showing it reliably references the knowledge base and safely handles requests. They also found that certain technical choices can affect performance, like maintaining Thai tone marks and enforcing retrieval for trustworthy answers.
retrieval-augmented generationThai traditional medicinelarge language model (LLM)dense-sparse retrievercross-encoder rerankingquery classificationzero-shot translationquantizationRecall@5knowledge base grounding
Authors
Pornthep Ukosaramig, Kobkrit Viriyayudhakorn
Abstract
We present TSWAP, a deployed eight-language conversational wellness advisor grounded, via retrieval-augmented generation, in a verified knowledge base of Thai traditional medicine and certified wellness providers. An unmodified open-weight LLM (Qwen3.6-35B-A3B on vLLM) is grounded on a ~30.6K-chunk Thai index by a hybrid dense-sparse retriever with cross-encoder reranking; a first-turn query classifier forces tool-based retrieval for entity lookups; a rule-based safety layer enforces medical scope and Thai emergency routing; and all eight languages are served zero-shot with translate-then-retrieve. We release the first Thai traditional-medicine/wellness retrieval benchmark (50 questions with gold document IDs; Recall@5 = 0.88), production QA logs (91.1% test-retest pass over 259 cases), and a 71-question frontier no-retrieval probe showing what each grounding pillar contributes: without the safety prompt the backend model family produced a full drug-dosing schedule and complied with out-of-scope requests, and without the knowledge base it produced zero verifiable provider recommendations. We further report two transferable deployment findings: English-calibrated 4-bit AWQ quantization corrupts Thai tone marks, and forced-retrieval routing is necessary for reliable grounding.