Manual data entry is automated via Optical Character Recognition (OCR). Invoices uploaded as PDFs are scanned by AI models to extract Supplier Name, Date, Invoice Number, and Total Amount, auto-populating purchase entries.
Automatic bank feeds download transactions directly. System match rules scan descriptions to match bank charge line items, checking offsets automatically against invoices.
手工数据录入现已通过光学字符识别 (OCR) 实现自动化。上传的 PDF 发票经过 AI 模型扫描，自动提取供应商名称、日期、发票号码及总金额，自动生成采购记账分录。
自动银行流水接口直接下载交易记录。系统对账规则扫描交易描述，自动匹配银行手续费等细项，并自动冲销匹配的发票款项。
Visual concept map explaining the key module topic: 解释该模块核心主题的视觉概念图：
Prompt: 'The OCR extracted an invoice total as RM 10,000 but the line items sum to RM 10,080 (including 8% tax). Write a rule prompt to identify and flag this tax invoice variance.'
Result: AI generates validation checks, flagging OCR scanning mismatches before ledger saving.
Prompt 提示词： “OCR 提取的发票总额为 RM 10,000，但单项金额之和为 RM 10,080（含8%税）。请编写一段规则提示词来识别并标记此税务发票的金额差异。”
效果： AI 生成校验规则，在入账前拦截并标记 OCR 扫描偏差。”
1. Which technology is used to automatically read text from scanned PDF invoices? 哪种技术常用于自动读取扫描的 PDF 发票中的文字？
2. Automatic bank feeds reconciliation benefits a business by: 自动银行流水对账对企业的主要好处是：
3. System matching rules commonly check bank descriptions to match: 系统对账匹配规则通常检查银行交易描述以匹配：
4. When a PDF invoice is successfully ingested by OCR, the system automatically: 当 PDF 发票被 OCR 成功采集识别后，系统会自动：
5. What is a common error in automated bank rules? 自动银行对账规则中常见的错误是？
Task 1: Write a regular expression to match Malaysian invoice numbers (e.g. INV-10022). 任务 1: 编写一个正则表达式来匹配马来西亚的发票号码格式（例如 INV-10022）。
Task 2: Which key transaction parameter is matched to reconcile accounts? 任务 2: 对账时匹配哪个关键交易参数来核对账目？
Task 3: SQL statement to count unprocessed OCR documents in table 'OCR_Docs'. 任务 3: SQL 查询：计算 'OCR_Docs' 表中未处理的 OCR 文档总数。
Match the term on the left with its definition on the right. 将左侧的术语与右侧的定义进行配对。