Supported Model
The supported model currently automates 25 document types using the latest AI, machine learning, and state-of-the-art language models to streamline document classification and data extraction. This automation ensures faster, more accurate sorting and distribution to the respective client entity.
By intelligently analysing the content, structure, and data of incoming documents, Artificial Intelligence (AI) powered by Large Language Models (LLMs) categorises documents based on predefined criteria such as text content analysis, pattern and layout recognition, document structure, and content relevance. To extract meaningful data, the system combines Natural Language Processing (NLP) for interpreting text, Optical Character Recognition (OCR) for digitising scanned content, and Machine Learning (ML) for identifying patterns and improving classification accuracy. LLMs enhance this process by enabling deeper semantic understanding, improving classification precision, and supporting advanced data extraction across diverse document types.
The solution extracts unstructured data points and converts them into structured data, which is then automatically mapped to the VMR lookup based on predefined rules. This structured data is subsequently applied as relevant tags, document names, and document dates. When the document is published from VMR to the Class Fund Documents DMS, the extracted structured data ensures accurate classification and supports seamless document automation.
Currently Supported Document Types
| Buy Confirmation | Listed |
| Sell Confirmation | Listed |
| Dividend & Interest Statements | Listed |
| Distribution Statement | Listed & MF |
| Dividend Reinvestment Plan | Listed |
| Tax Statement | Listed & MF |
| Managed Funds Redemptions & Exit Statement | MF |
| Cash Statement | Bank |
| ATO ABN Notification | ATO |
| ATO TFN Notification | ATO |
| ASIC Company Statement | ASIC |
| Real Estate Income Expense Statement | Property |
| Water Rates | Property |
| Council Rates | Property |
| Strata Levy | Property |
| ATO Commutation Authority | ATO |
| Notice of Intent to Claim Personal Super Contributions | ATO |
| SMSF Deed of Amendment/Variation | Compliance |
| SMSF Deed of Appointment/Retirement | Compliance |
| Investment Strategy | Compliance |
| ATO Trustee Declaration | ATO |
| Request to Adjust Concessional Contributions | ATO |
| Pension Commutation | Compliance |
| Member Insurance Policy | Compliance |
| Holding/Transaction Statement | ASX, APIR and Unlisted |
Unsupported Model
Unsupported document types in VMR refer to documents that are not currently automated for classification and data extraction. At present, a total of 76 document types remains unsupported. The Human-in-the-Loop (HITL) process enables users to search for these document types within VMR, manually initiate account mapping, and then publish the file and associated data into Class Fund Documents DMS.
Documents received in the solution that fall outside the supported model are initially classified as “Other Statement.” The HITL process allows users to open the source document, update the document type from “Other Statement” to the correct category, and manually map data points using the VMR lookup. Once mapped, the document is placed in the “READY” queue for publishing to Class Fund Documents DMS.