Having AI read the CIMs for you would speed things up like crazy.
I know they're confidential documents, but what if the AI platform doesn't retain, reveal or use the data for training ever? They're very clear on this for their website. For example, microsoft azure's wrapper on GPT4 claims this.
From my understanding, if the AI is SOC2 compliant, that they are considered "secure" for any private, confidential data.
Anyone familiar with the legal aspect here, or have an idea if this is possible?
I think the easiest way to CYA would be to run Llama3 with a PDF/OCR reader. This guy seems to have it figured out. https://medium.com/@recursechat/local-docs-local-ai-chat-with-pdf-locally-using-llama-3-a3b9a3a0c055
You could use a tool like jan ai to run the LLM locally pretty easily (DM me if you need help executing this, have done it for other tasks).
You'll still need to do some prompting to avoid classic LLM hallucination issues and math errors. I'd also advise chunking the CIM into parts to upload, as LLM performance degrades if you try to upload too much
Facere dolor dolor nihil consequatur sint. Quod aperiam reprehenderit eos dicta dicta sed. Tempore quis reprehenderit iure. Est eaque exercitationem hic ad vitae dolorem. Ab pariatur fugit odit omnis id consequatur. Repellat corrupti ratione distinctio ut voluptas voluptatem.
In ut minus adipisci ut est pariatur. Eveniet quas ut itaque impedit adipisci corrupti est occaecati. Ut quia natus nostrum nostrum facere non.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...