This is built on the latest versions of ChatGPT Assistants API. We use blend of models depending on the stock (gpt-4-1106-preview, gpt-4-32k, and gpt-3.5-turbo).
Yes, you could. You would be limited to 180 days of history on initial pull and your data would always be up to 1 day behind. Otherwise, yes, you could. It doesn't really feel like this product is aimed towards aggregators or dashboard and more for underwriting and bank account validation.
The limit to only daily pulls and up to 180 days of historical data is pretty disappointing. Would expect Stripe to push the envelope here and move down to near-instant updates and full historical data. This is basically a knockoff of existing solutions done at par or worse which is surprising to see from a company like Stripe. Maybe they've lost a bit of their magic or focus. Will be interesting to see how everyone adapts and improves to this announcement.
since they are heavily based on existing aggregators, they are limited by the lowest common denominator (if a good chunk of e.g. Yodlee-scraped credit unions only support 180 days, then that's a limiting factor for everyone).
I would expect they will gradually improve and cut out other aggregators and replace with direct connections that have more & better data as they go along.
Or maybe not, because with Oauth they then become at the mercy of the bank that issues the token (banks can revoke tokens or attach individual conditions on what data is accessed and how much).
Yep. In addition to being toxic and wrong and pointless, it also reveals that OP isn’t actually very good at the non-JS parts of front-end engineering. It’s easy to assume that something is basic when you only have a basic understanding of it.
I work with people and keep pushing to ship it but all the non-engineer members are pushing for more features. I don't think it's just a software developer issue.
My point was many software engineers see the product in their mind and think "I can build that!" and think that building it is 99% of the work and if they just built it the money will flow. But it turns out building it is only 10% of the work and when the other 90% of the work presents itself the drive to ship disappears.
In the OP's case the people that shipped found a way to get all the parts needed to ship read. The person didn't ship probably just wrote some code and never even considered the other 9 of 10 things that would need to be done to actually ship.
I'm probably thinking of myself to some degree. I made a product. Maybe I didn't have enough confidence to find partners or investors. I piddled with it for about 18 months, had a few public displays, and then at a trade show I saw someone who was serious with basically the same product but unlike me they found investors, hired people, and had a real company. They didn't just make some tech and a demo (what I did).
Happy to answer any questions!