My favourite part is probably the diverse and flexible amount of programs that Carbon is able to provide us to train our model. For instance, at first, we were using the common webscraping tools for GDrive and OneDrive, but eventually, we unexpectedly had some applications with Notion and Slack as well, but Carbon already had those set up when we needed them. If we have new programs, we usually just send them a message and they are probably happy to set that up for the other companies as well.

Our favourite part about Carbon was probably how convenient their solution is. When creating our LLM, we tried creating a data connector system ourselves, but it’s a bit more tricky than we had initially thought, mainly because of the sheer amount of programs we would have to set up to be compatible with (not to mention it’s probably going to change over time), and that we were a relatively new startup and didn’t have much cash to spare. But when they reached out to us, their product was very streamlined since it integrated all of those programs that we used into one common API and all we needed to do was have our product manager use their portal and spend a week on how to use all of the features with the help of their support team and our developer.

Something that I think really distinguishes Carbon from the other companies out there is that they will customize their programs to your requests. Each time we messaged them about the feedback we have received from our users, their team always adjusts accordingly and develop new API and UX improvements that improve our business and our customer experience. I feel that even if we suggested for them to develop an integration for a new program, they would probably do it within a few weeks.