potai.com

Cannabis trust infrastructure

A working thesis. Inviting comment from operators, suppliers, CPAs, and anyone who has lived the problem.

The problem

Cannabis is a cash-heavy, federally illegal, state-regulated industry where nobody can easily prove what happened. That single fact creates most of the operational pain you already feel.

Banks discount reported revenue because they cannot verify it. Lenders charge punitive rates or refuse to underwrite at all. Landlords demand personal guarantees on commercial leases. Suppliers extend net-30 terms that quietly become net-90, with no shared memory across the industry to flag retailers who pay one supplier on time and stiff three others. Form 8300 cash reporting obligations get handled in spreadsheets the night before deadlines, with the federal exposure that implies.

None of this is a marketing problem. It is a verification problem.

The thesis

PotAI would be a neutral verification layer for cannabis businesses. Not a bank. Not a lender. Not a payment processor. A trust infrastructure company that turns cash activity, wholesale invoices, and trade-payment behavior into auditable, permissioned records that operators, suppliers, and their counterparties can actually rely on.

The bet is that the industry's missing layer is not more capital or more software. It is shared, verified memory.

Two product wedges

Wedge one

Form 8300 compliance

The IRS requires any trade or business receiving more than $10,000 in cash in one transaction or related transactions to file Form 8300, generally within 15 days. Marijuana-related businesses are explicitly named in IRS guidance.

Most operators handle this in spreadsheets. PotAI would detect reportable cash events from POS exports and cash close records, prepare filing packets, track deadlines, generate the year-end customer notices required by January 31, and store the supporting audit trail.

Wedge two

Trade credit ledger

Cannabis suppliers extend net terms to retailers with no shared memory across the industry. A retailer who quietly turns net-30 into net-90 can do it to ten suppliers in sequence before anyone connects the dots.

PotAI would let suppliers contribute A/R history into a permissioned network. Descriptive facts only. Paid on time. Paid 31 to 60 days late. In dispute. Written off. Retailers get notice and a real dispute and correction workflow before any negative event is shared. No risk scores. No suggested credit limits. Not yet.

What this is not

PotAI would not move money, hold balances, or route payments. That keeps the first version of the company out of the hardest licensing zones and lets it stay focused on what it is actually good at, which is verification.

It would not sell credit scores or suggested credit limits at launch. Once you publish a number that looks like a score, every adverse action becomes a potential dispute about how that number was derived. Descriptive facts about events that actually happened are a much sounder place to start.

It would not be a blacklist. Retailers would receive notice before any payment event was shared with the network and would have a real correction and dispute process. A sloppy version of this product would be worse than no product at all.

Where this is in the process

This page is a thesis, not a launch. The next step is a small validation sprint with hard signals, not survey interest.

For operators, the test is whether anyone will pay for a 12-month Form 8300 cleanup and filing packet review. For suppliers, the test is whether anyone will sign a data contribution agreement and share current A/R aging for normalization and slow-pay analysis.

I am looking for three operators and three suppliers who are close enough to the pain to help shape the first version.

Tell me where this is wrong

Especially if you are one of the following:

Useful prompts if you are not sure what to write:

Reach me directly