What data for this category looks like
Data for this category comes from enterprise ITSM systems, phone or chat logs, asset systems, and internal knowledge bases. Commonly processed documents include service request tickets, incident tickets, historical processing records, and user satisfaction feedback forms. Data update frequency is daily or real-time. Fields include unique ticket identifier, submitter information, problem description, processing status, associated asset number, satisfaction score, and more. Some data comes from structured forms, while some comes from unstructured chat texts and handwritten records, with no unified fixed format.
What constraints these characteristics impose on the document parsing and chunking link
Since data sources include structured forms and unstructured text, the parsing link must support both tabular data and free text. Ticket documents contain a large number of short text snippets and long description paragraphs. Chunking must avoid breaking the connection between key problem context and processing steps. Real-time updated data requires the parsing process to have low latency characteristics. Differences in fields across multiple sources require the parser to support custom field mapping. Some tickets contain sensitive information. Chunking must ensure that sensitive fields are not split into unrelated chunks, while meeting data processing compliance requirements.
How to set the configuration
This document is written based on FastGPT v4.17.0.
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
chunkSize | 800–1200 characters | Adapts to the average length of ticket text, avoiding breaking the context of problem descriptions and processing steps |
chunkSplitter | Paragraph + semantic splitting | Compatible with structured ticket forms and unstructured chat texts, retaining semantic integrity |
UPLOAD_FILE_MAX_SIZE | 100 MB | Adapts to the common size of ticket attachments on enterprise IT service desks, preventing upload errors |
PARSE_FILE_TIMEOUT_SECONDS | 60 seconds | Matches the low-latency requirements of real-time ticket processing, preventing parsing interruptions caused by timeouts |
customPdfParse | Enabled | Supports parsing PDF-format service manuals, historical ticket documents, and compliance files related to tickets |
chunkSettingMode | Custom mode | Allows adjustment of chunking rules for different field types of tickets, adapting to multi-source data structures |
Three common errors
- A 413 error occurs when uploading ticket attachments. The cause is that the
UPLOAD_FILE_MAX_SIZEconfiguration has not been adjusted, exceeding the file upload limit allowed by the platform. - Only the first two columns of content are read when importing CSV-format ticket data. The cause is that custom field mapping has not been enabled, and the default parsing logic only reads the first two columns.
- The external pdf-maker parser fails to work when used. The cause is that the corresponding parsing key parameter has not been correctly filled in the system configuration.
How to confirm the configuration is complete
- Upload a test ticket PDF attachment, check whether the upload progress and parsing results are normal, with no 413 error.
- Upload CSV-format ticket test data, confirm that all preset fields are correctly read, and the situation of only reading the first two columns does not occur.
- After configuring
customPdfParse, submit the external pdf-maker parsing key, and verify that the PDF document parsing process is triggered normally. - Adjust
chunkSizeto different ranges, parse long-text ticket documents, and check whether the chunking results retain context integrity.
The values provided are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-10-19.