When our customers were using our chatbot, they were only able to upload local documents, which were constrained by length. Additionally, those important materials were not saved after the chat session ended for technical and security reasons. But Carbon was able to be our solution since they converted those into vector embeddings and stored it in a database. We were able to have users web-scrape and upload local files, create custom tags to their documents, and implement a fast semantic search to retrieve content from uploaded documents.

We probably have to say that we love their approach towards helping their customers the most. We first reached out to Carbon after our customers had requested the option of uploading files as part of their AI chatbot that we delivered as part of our programs. It was difficult for us to implement ourselves due to its complexity, but it was a smooth effective resolution with Carbon’s support. After successfully implementing that with Carbon, we wanted to add our pre-built data connectors next, and they leveraged our white-label service and custom branding across all user-facing components, launching Google Drive and Dropbox connectors just a couple weeks later.

At our company, we mainly used Carbon for connecting some applications/programs to train the LLM that we were providing our clients. But for us, we like their product’s scalability and reliability the most. They offer auto-scaling and 99.995% SLAs, which meant we can handle the increasing demand without compromising reliability or performance.