Open bitcoinbrisbane opened 9 months ago
User Stories for the Document Management System MVP
User Story 1: Document Addition and Metadata Extraction As a law firm clerk, I want to add PDF documents to the admin site and have their metadata automatically extracted, so that I can efficiently organize and process these documents for Full and Frank disclosure requests.
User Story 2: Request Table Management As a law firm clerk, I want to input document request details into a Request Table, so that I can keep track of various documents requested, their metadata, and associated deadlines.
User Story 3: Document Organization and Indexing As a law firm clerk, I want the system to automatically organize documents into a file structure and rename them, as well as create an index or table of contents for the recipient, to streamline the preparation of documents for the other side.
User Story 4: Automated Email Generation As a law firm clerk, I want the system to generate emails complying with the other side's instructions when documents are ready, to reduce manual effort and errors in communication.
User Story 5: One-time Document Request Links As a law firm clerk, I want to send a one-time link to clients requesting documents, to ensure secure and specific document submission.
User Story 6: Client Document Submission Interface As a client of a law firm, I want an easy interface to submit the requested documents through a one-time link, to ensure I can comply with the document requests efficiently and securely.
User Story 7: Automated Document Validation As a law firm clerk, I want the system to use Local Chat GPT to assert the correctness of the client's documents, checking for date ranges, bank, and account names, etc., to ensure the accuracy and relevancy of the documents received.
User Story 8: Automated Client Communication for Future Implementations As a law firm clerk, I want the admin page to automatically draft and send emails with one-time links to clients, to further streamline the document request process.
Prompt for "created" field
Prompt for "deadline" field
Problem
The “other side” asks for documents to be supplied for Full and Frank disclosure. Collating, reviewing and processing these files and creating an index for the other side is time consuming and error prone.
Overview
Step 1
The firm adds the PDF to the admin site and meta data is abstracted. Add the letter meta data to the Request Table. The fields can be prefilled by ChatGPT.
Request Table Schemea
Endpoints
GET /document POST /document/request
Eg
Step 2
Add the document requests to the table and reference the request. A view can be created if the request is completed, and ready for inspection.
Step 3
The client is sent a one time link requesting documents, such as https://doc.firm.legal/8fd3c03a-6f87-4bc3-aaa0-5c49e76e9b6f. For now, this is a cut and paste by the clerk.
Step 4
Upon receiving the one time link, the SPA will present the required documents form. Note, this can be one to many. The client may or may not see the original document. This will be a switch and default to "show".
Step 5
Our API will use Local Chat GPT to assert the correctness of the clients documents submitted, looking for dates ranges, bank and account names etc.
Step 6
The API will sync back to the firms cloud storage under some folder structure. This will be configured per firm.
Step 7 (future)
The admin page will automatically draft an email to the client and send the one time link from Step 3.
Future
It would also be good to have the client sync their personal drop box, google drive, one drive account ect. They can make folders per request and bi directional sync.