Show HN: I built a forensic accounting tool for high-conflict litigation(exitprotocols.com)
exitprotocols.com
Show HN: I built a forensic accounting tool for high-conflict litigation
https://exitprotocols.com/
1 comments
OP here.
Just to share a bit of the engineering headache behind this: The hardest part wasn't the Django backend, it was getting the LIBR (Lowest Intermediate Balance Rule) logic to handle edge cases in the PDF parsing.
Bank statements often group transactions by date, not time. If a user deposits $5k and withdraws $5k on the same day, the order matters for the 'dip' calculation. I ended up having to write a heuristic that forces 'Withdrawals First' (worst-case scenario for the claimant) to ensure the report stands up to conservative judicial scrutiny.
If anyone here has worked with financial event sourcing for legal compliance, I'd love to know how you handle same-day timestamp ambiguity.
Just to share a bit of the engineering headache behind this: The hardest part wasn't the Django backend, it was getting the LIBR (Lowest Intermediate Balance Rule) logic to handle edge cases in the PDF parsing.
Bank statements often group transactions by date, not time. If a user deposits $5k and withdraws $5k on the same day, the order matters for the 'dip' calculation. I ended up having to write a heuristic that forces 'Withdrawals First' (worst-case scenario for the claimant) to ensure the report stands up to conservative judicial scrutiny.
If anyone here has worked with financial event sourcing for legal compliance, I'd love to know how you handle same-day timestamp ambiguity.
I’m a solo developer who realized that in high-conflict divorce cases, the party with the most organized data wins. Forensic accountants charge $400/hour to trace assets, which makes financial justice inaccessible for many.
I built Exit Protocol to automate this. It’s a forensic intelligence tool that ingests PDF bank statements and uses the Lowest Intermediate Balance Rule (LIBR) (See v. See, 1966) to mathematically separate marital funds from separate property.
My very first post: https://news.ycombinator.com/item?id=46350044
The "Defense-Grade" Features (v2 Update):Realizing my users are often under digital surveillance by a spouse, I overhauled the security model to be "adversarial-ready":
-Protocol 0 (Anti-Coercion): I implemented a "Duress Password." If a user is forced to unlock their device, entering this secondary password loads a fully functional "Decoy Dashboard" (a generic budgeting app) with fake data, hiding the forensic evidence.
-Deposition Killer: A feature that cross-references financial transactions with text message timestamps to find contradictions (e.g., User texts "I have no money" on Tuesday -> System flags a $5k Casino withdrawal on Wednesday).
-Sovereign Architecture: For law firms, the system supports a "Bring Your Own Key" (BYOK) model where the entire stack runs in an air-gapped Docker container.
Stack:
-Backend: Django 5.0 (Monolith)
-Queue: Celery + Redis (for parsing 1,000+ page PDF discoveries)
-Logic: Custom Python implementation of LIBR tracing algorithms
-Security: Fernet (AES-256) encryption at rest
Try the Simulation (No Login Required): I know HN hates sign-up walls, so I built a "Live Tactical Simulation" (link in the footer) that pre-loads a fake case ("Operation Sterling") so you can play with the Asset Graph and LIBR engine immediately without registering.
I’d love feedback on the Duress implementation specifically. Is a decoy dashboard enough, or should it trigger a silent nuke of the local session storage too?