From financial documents to a reviewable divorce case file.
Divorce Masters is a legal technology platform for attorneys and their teams working through financial disclosure. The job spans clients, documents, missing information, financial review, and reports. A useful system has to keep those pieces connected to the same matter.
We built the product around that workflow: a workspace for the firm, a document-upload portal for the client, and AI-assisted review that leaves the legal team in control. The engagement covered the application and its backend, not just an AI interface placed over an existing process.
The product serves people doing different parts of the same job. A client needs to understand what to provide. A paralegal needs to see what is missing and organize what arrived. An attorney needs to inspect the financial picture and decide what it means. The build connects those roles without treating document collection, AI output, and legal judgment as the same task.
| Project context | Public detail |
|---|---|
| Product | Divorce Masters — financial disclosure software for divorce attorneys |
| Primary users | Attorneys, paralegals, firm administrators, and clients |
| Unit of work | The matter: its documents, disclosure checklist, review, and reporting |
| Collection framework | Florida Rule 12.285 mandatory disclosure |
| Delivery scope | Firm workspace, client portal, AI-assisted review, and financial reporting |
| Implementation record | July–August 2026 product and financial-workflow delivery |
Financial disclosure is not complete when a folder contains files. The team needs to know which records were requested, what arrived, what still needs attention, and which financial findings require a human decision. Clients and staff approach that work with very different levels of context.
The product brief made the constraints explicit: clients may be on a phone, under stress, and using the portal for the first time. Attorneys and paralegals move between many matters. The design needed a clear next step for the client and a precise review surface for the firm.
Financial categories also need more detail than a broad AI label can provide. A transaction marked utilities does not, by itself, tell the team whether it belongs under electricity, water, telephone, or gas on a financial affidavit. The workflow has to support that distinction rather than hide it inside a total.
There is a second distinction inside the checklist: a missing upload is not necessarily a document that exists and has not been sent. The portal supports an explicit attestation that a requested record does not exist. Giving that situation its own state is more useful than asking staff to infer the answer from an empty document slot.
Clients see the documents requested for their matter and upload against the Rule 12.285 checklist. A distinct attestation lets them indicate that a requested record does not exist, rather than leaving the team to interpret an unexplained gap. Collection is tied to the request and the matter, not merely to a folder of incoming files.
Attorneys and staff organize matters, inspect documents, and review AI-flagged financial issues. The matter stays the organizing unit, so review work remains connected to the case it belongs to. That gives the client-upload process a destination inside the firm's working file instead of making it a separate intake tool that ends at receipt.
The transaction ledger supports filtering and paging so staff can work through a targeted review queue. Corrections distinguish broad spending categories from the finer affidavit lines: utilities alone cannot identify an electricity or water expense. Matter-specific rules take precedence over firm-wide merchant defaults, so a reusable convention does not override a decision made for an individual case.
The reporting backend carries affidavit-line detail into the workbook alongside the spending summaries. The finer breakdown is kept separate from the category totals so the same dollar is not counted twice. Staff can work with the broad spending picture and the more specific reporting detail without treating those two views as additional transactions.
The application uses Next.js for the firm workspace and client portal, with a NestJS and Prisma backend and Supabase authentication. Tenant-data requests go through the backend using the caller's own token; the frontend does not hold service-role credentials.
The financial-review implementation separates AI-proposed transaction categories from staff corrections and affidavit-line assignments. Firm-level rules can provide reusable merchant defaults, while a rule specific to a matter wins when that case needs different treatment.
That separation also shapes how the system improves. A firm-wide merchant default can provide a starting point for repeated review work, but a case-specific rule has higher priority. The reusable part is a categorization convention, not a transfer of one client's financial records into another client's matter.
The July 2026 implementation established the branded workspace, client portal, and disclosure-checklist flow. August delivery added filtered transaction review and the finer affidavit-line reporting work. These are implementation milestones: a history of what was built, not a claim about how many firms adopted it or when a particular practice went live.
Based on Divorce Masters product documentation and July–August 2026 implementation records. No measured time-saving, accuracy, or adoption figures are claimed. Client documents and matter details are not reproduced here.
The client-facing side gives people a concrete task: provide the records requested for their matter. The firm-facing side brings the checklist, documents, flagged issues, and financial corrections into the same product.
The useful handoff is not an AI answer presented as a finished legal conclusion. It is a document, a flagged issue, or a financial line in a place where staff can inspect it and do the next piece of work. The workflow makes room for corrections instead of treating the first extraction as final.
The intent is to put more of the legal team's attention on exceptions and judgment, rather than on assembling the working file. AI assists with organization and review; it does not replace the attorney's responsibility for the conclusions or the material used in a proceeding.
The work is not to make a legal decision disappear. It is to make the information behind that decision easier to collect, inspect, and correct.