I started building something like this for myself and though it would be useful. I have a decent mac only working working but what you have built is great.
This one is completely open source and local, no api calls required at all. Quality can better with newer models which has to be worked on.
I'm building AI for documentation. It will answers users queries related to your product through videos, documentations and blogs. Reducing the load on customer support.
Here's the summary:
- read all your sources
- public websites, docs, video
- answer questions with confidence score and no hallucinations with citations
- cut support time and even integrates directly into your customer facing chatbots like Intercom.
Still deliberating on the business model. If anyone would be interested in taking a look, I would love to show you.
I am interested to take a look but some answers before it might be great?
One of the issue we are facing is the upto date documentation. Like you have document A with information on Doc A, but now there is a document A.1 is written which has the updated information on Doc A.
Yes we have thought about that and we have multiple solutions to that.
- webbhook triggered - when a document is updated some CMS/tool provide webhooks triggering capability, which you can use to reindex that page
- time based triggers - you can set a time like a cron and the document will be scanned in that time and checked if something has changed it will be reindexed
Here's the summary:
- read all your sources - public websites, docs, video
- answer questions with confidence score and no hallucinations with citations
- cut support time and even integrates directly into your customer facing chatbots like Intercom
Still deliberating on the business model. If anyone would be interested in taking a look, I would love to show you.
I'm not an expert on disaster recovery, but have to think of it for $day_job on occasion. The way i think about it is something along the lines of when (not if), disaster strikes, what needs to be saved vs what can wait a while to have it recovered...plus what level of effort are the different sets of activities worth. For example, if i host my public blog website on my machine at my home, and power goes out (thus making my blog unavailable during the power outage), am i ok with this, or do i need to ensure that it can be brought back up? If i feel that it needs to be brought back up during such a scenario, how much effort (which really means time plus money) do i wish to plan for? Obviously this is a vast oversimplification...but that kind of thinking has helped me bring up topics and questions in advance in order to plan for extreme scenarios...and you know what? Sometimes the things people think they "really need to be up/available" are not that important. Other times, sure, but then you commit, plan, test, and prepare for such truly essential activities. (And, of course, other times, whether the thing that needs to be recovered is important or not, you might be compelled by laws, regulations, contracts, etc...but that's a whole separate topic.)
So, again, start at the end - when disaster strikes - and consider what you think is worth keeping/bringing back up, and then plan backwards from there to make it so.
Agreed. I've been using it for all DNS and CDN for over two years now. Great company, great support, great performance, great API. Everything. Love it. I'm a big fan.
