What this category's data looks like
In the enrollment consultation conversion scenario at training institutions, materials for multilingual materials and terminology processing mainly come from three categories: enrollment brochures, consultation records, and trial class schedules. Enrollment brochures provide course names, fee definitions, start dates, and enrollment terms; consultation records carry the consultant's name, intended course, follow-up date, and advisor signature; trial class schedules record the trial time, instructor, classroom, and confirmation receipt. The three materials play different roles when entering the model integration and configuration stage: enrollment brochures and trial class schedules serve as terminology sources and reference bases, while consultation records serve as real-world conversation samples for how terminology is applied. The material update rhythm usually follows enrollment batches and trial schedules, and the field characteristics are mixed languages, the same term written differently across materials, and dates and quantities that need to be aligned item by item. The output of the glossary and comparison draft depends on these materials first completing field-level verification, then entering the dialogue and retrieval stages to perform terminology consistency checks.
Configuration and implementation checks
This step is executed by the server-side model channel configuration, outputting model binding relationships and channel responses. The "unified glossary and comparison draft" is then completed by subsequent AI nodes, code nodes, or manual work. Models are divided into five categories: language, indexing, reranking, speech synthesis, and speech recognition. Different stages bind different model types, and the configuration side is only responsible for matching types to stages and confirming the model identifier returned by the channel.
| Object | Check | Verification action |
|---|---|---|
| Enrollment brochure | Course names, fee definitions, start dates, and enrollment terms must be aligned item by item with consultation records; the same term must be written consistently throughout, and the translated name must be traceable back to the source text | In the model configuration, check the type labels of the language model and indexing model, compare them with the model names returned by the channel, and confirm that brochure terms can be hit by the same identifier on the retrieval side |
| Consultation record | The consultant's name, intended course, follow-up date, and advisor signature must be consistent with the trial class schedule; missing signature or receipt is regarded as not closed-loop | Check the channel configuration and key source, and confirm that after the dialogue side is bound to the language model, the terminology in the consultation record keeps the same wording in the response |
| Trial class schedule | The trial time, instructor, classroom, and confirmation receipt must match the dates and quantities in the enrollment brochure; when versions are not synchronized, the latest batch prevails | In the model configuration, check the type label of the reranking model, and confirm that after retrieval results are reranked, the reference text tokens and array item count of the terminology comparison draft are recorded separately |
Three common errors
- Symptom: The same course name is written differently in the enrollment brochure and consultation record. Cause: The glossary is not unified, the two materials are maintained separately, and versions are not synchronized.
- Symptom: The dates in the trial class schedule do not match the start dates in the enrollment brochure. Cause: After the schedule was updated, only one material was changed, while the other still uses the old batch.
- Symptom: The consultation record lacks an advisor signature or receipt. Cause: The follow-up process is not closed-loop, and field-level verification was not performed before the materials entered terminology processing.
How to verify the setup
Acceptance for this feature: run one dialogue and one retrieval, check the model identifier in the response, and confirm that the values of maxContext, maxTokens, and quoteMaxToken in the model configuration are consistent with the documentation.
Business test: sample three terms and check whether their wording is consistent throughout, and confirm that each translated name can be traced back to the source text.
- In the model configuration, check item by item the type labels of language, indexing, reranking, speech synthesis, and speech recognition, and confirm that they correspond to the stages to be used.
- Check the channel configuration and key source, and compare whether the model names returned by the channel are consistent with the names listed in the documentation.