They're provide a tech + manual bookkeeping service, where you have an actual accountant to talk to, aided by technology.
I use them for my LLC, and it's been pretty paperless so far. They can link directly to all my accounts (Chase, US Bank, Citi) for ongoing accounting work. For historical or one-off statements, I upload a pdf statement with highlighted items and notes (fairly simple to do in Mac Preview).
Any questions they or I have is communicated through their dashboard, including assumptions they're making and categorizing outstanding charges.
If so, did they say they could satisfy those requirements, and then later fail to do so? If so, did they give any reason why they misled you?
They did not maliciously mislead us, but failed on their promise because apparently the ‘automagical’ part does not seem to work out so well. Yet, that, imho, is more the fault of the implementation and of its users (the accountants), than it is because such things are too hard to accomplish with the state of current software development (and ‘AI’).
For one, I found it beyond comprehension that any such software would blindly rasterize well-structured PDFs, as does the software used by our accountant, to then OCR the resulting low-res jpg, while wastefully of resources throwing away most of the relevant information, with obviously bad ‘detection’ results. Worse, duplicate uploads would differ, resulting in double bookings requiring the accountant’s assistant using the software to manually fix those errors produced by the software. In the end we end up with a much more expensive accounting bill than we would have had when we sticked to a paper-based flow.
But is doable, and I truly hope some startup grabs the opportunity to implement it well.