1. There will never be a viable solution until an ISO standard is published and governments force organizations to use accounting systems certified to follow the standard. 2. Vendors would have to open-source the code that allows us to understand what happened to the data after it was entered into fields and how it's processed. 3. Accountants are simply used as pawns when someone needs to go to prison for embezzlement. Accountant could never prove if the data in the report is interpreted correctly. I built such system, no one wants to use it because it's too different, has no support, no wide adoption. People always want something more mature. They don't realize all these vendors are making up their own rules as they go. Intuit, SAP, they're all just protection against audit. Let's say you have line item called "muffin" and it was sold for $80. Maybe it's a valid operation, maybe it's a sign of embezzlement. The point is, there is no way to check the largest organization because, say, YouTube, has so many operations, you'd need to use LLMs to check each one automatically against possible examples of questionable operations, and that's after you have standard schema, which none exists today. You're best waiting until modern Luca Pacioli is born and publishes that standard and everyone agrees to follow. So, best you can do is wait. Simply no one cares about how to do it the right way. Everyone has their own interpretation of what accounting system schema should look like. The solution is so obvious that if implemented people would end up in prison because they just couldn't follow the rules most likely going forward. They'd be exposed as only being in existence because the government doesn't have the audit capacity. Again, YouTube, if you had to check legitimacy of transactions manually, or even after triggering anomaly, you'd still need five seconds on each transaction at minimum. Unless it is acceptable to justify that some transactions will never be checked for legitimacy. Every time you watch YouTube video, you are shown an ad. That ad is associated with an amount someone was willing to pay to get you to watch it. There are simply so many entries in the journal every second, that if you had to spend five seconds on each transaction to verify, you couldn't because the number of transactions in the world is increasing faster than resources we have to check them. Especially in high frequency trading, again, if the journal has to record a debit and credit, that's two rows of data in the database. We have no audit capacity. Organizations know this and simply know which numbers on tax forms don't trigger audit. That's why tax forms often have occupation, EIN, and historical anomalies in income/expense. As long as you don't trigger those, you will never be checked. Organizations know this. The accounting department is simply smoke and mirrors and they'll pick an accountant who can't prove the system misinterpreted or didn't correctly save or calculate, and they'll end up in prison with no ability to prove their innocence unless the source code is released. Furthermore, quickbooks and others don't let you receive payment against individual line item on the invoice, which is important if your invoice has multiple vendors you are a funnel for and they have different terms where partial payments have to be allocated correctly. Anyone entering this space has no idea what they want and what they're getting themselves into. You are better off waiting for Pacioli V2 while using QB than anything. It benefits oneself to have perfect accounting system, but it's an insane effort, you literally become accounting system vendor instead of your core business. You'll be so good at providing such software others will envy you and want to use your system. Besides, any effort must be international. You should be able to use QB in North Korea, but you can't. Which means QB is not that good. When Pacioli published his double entry accounting method, he didn't have computers, databases, and software. There was no ISO 500+ years ago. If Pacioli published something today, it wouldn't by a system, but a standard accounting system would follow. It's already widely understood, there just needs to be ironed out some details involving transaction scope which Pacioli didn't have that guarantees entry of all or none of records, which is kinda important when you treat double entry as one. The standard schema is actually simply, it's just your basic debit/credit with normalization via many-many relationship. You simply have to link two or more journal entries with whatever they represent, invoice, payment, reconciliation, inventory movement. Then you apply convention based naming and you have a database when ejected can be audited and checked almost automatically. This wasn't exactly possible before LLMs, the structured output in LLMs helps with this as it naturally should. Again, if we did this, then everyone would be playing by the same rules and the world would probably flip upside down, trump would end up in prison and it would be end to putin. Anyone that doesn't understand this is at disadvantage. Imagine if you had custom accounting system and everyone in your organization died, does that mean the organization can't be audited to pay its creditors or continue existence under new management? Remember, all accounting does is record expenses and income most of the time. YouTube records theirs automatically, which is another misunderstanding. They probably have two ledgers, one for income and another for expense. There's another misunderstanding, there is no need for ledgers anymore, you just need one timeline of financial events in such way that you can answer any question without involving the person originally responsible for the operational transaction. What might happen is everyone will end up doing their own accounting in the end. The only way intuit can survive having to implement such standard, depends on if they're smart enough to only open-source the code that's responsible, the entire stack, for making invoice, payment, other transactions. You'd need to dump that information into journal records because code could change and you'd need to prove that previous transactions were recorded differently maybe and at what point the change happened. Either QB code went through changes that make it act different after 26 years, or it's never changed, which means there is no way QB got it right the first time. We have companies like SAP too, what are they doing different? Russia, China, have their own systems for their markets. This is extremely serious, but no one will do anything about it because it would make everyone auditable. Then the only question why we aren't auditing them is a click away, and you wait. Besides, what's the threshold, .01, .05? When is the bulk transaction sufficiently different to where you need a human to verify it's legitimacy. If sales team records invoice at $1000 and then changes the price to $500, that should trigger a notification to the boss, does QB do that? Can it? Can SAP? Maybe, but it's more complicated than just understanding that all you need is a system to model financial operations which are just rows of data in SQL database wrapped by convention and linked to whatever they are for. It's just impossible. I've been going at this for 3+ years. No one understand this and they wouldn't want to because all businesses would be found in serious violations. They only survive because we can't audit them because we have no capacity to. There needs to be a speed limit to how fast journal entries are recorded can be only as fast as they can be checked by an auditor. If we simply say that an organization is too big to audit, then we fail. No one can compete with that. Which is why you will be stuck with garbage like intuit or have to roll your own but they again you have to support it which pretty much makes you accounting system vendor, so why are you in whatever business you are in if creating your own system would put you into competition with QB, SAP, etc.? It's just impossible. But on the other hand, once you get Pacioli to publish a standard there will be fireworks. There is no way high speed trading would be allowed to exist. There is no way YouTube could undergo an audit. We won't know until YouTube is audited mechanically, which is only possible with a standard. People don't remember, but we used to keep paper journals before computers. Then computers came out and it's gone to siht. You think Pacioli, if he was around today would approve of QB? Or, since, Pacioli double-entry accounting method lasted 500+ years up until 30-40 years ago, he would have created something like QB today? He'd tell QB to follow a standard. Those ISO lunatics are from Geneva, that's the last place that would approve an international standard like that. We're screwed until a big rock from space takes them out.
almost forgot, don't forget about disaster recovery, that's the whole point of these system, so you have reliable access to information with least obstructions.