Jeff Schneider Transforms Enterprise Chat with Imprompt's GPT Integration
Jeff Schneider, Founder and CEO at Imprompt
Founder and CEO at Imprompt
Jeff Schneider is the founder and CEO of Imprompt. Jeff started his career as a software engineer, and authored the first book on Enterprise Java. Later, he founded MomentumSI, an API and cloud consulting company which was acquired by VMware, where he led VMware's cloud professional services in the Americas.
Transcript
Everyone wants to chat to GPT experience but the mega vendors often try and give it to you in one form factor. Imprompt provides a chat to GPT experience to large enterprises so that they can customize it and bring their own solutions to the challenge that we're trying to solve for customers is really empower them to be able to use the new generative AI on their own terms. We want them to be able to use all their existing IT infrastructure and to bring it into the GPT world and not have to worry about the needs of the mega-vendors.
When we started the journey, we looked at Pinecone and Weaviate and some of the other startups in the space. Unfortunately, we did sign with one of them. But we found that they dropped our database unintentionally for unable to recover for six days. With that we decided that it was more important to find a vendor who was more reliable and knew how to move fast. DataStax was the partner because they thought like a professional journeyman, but they were also able to move like a fast startup. We're always concerned about bringing product to market rapidly. And we did so with the DataStax product in very short order.
When we were architecting the product we contemplated how we are going to deal with files generally, a PDF is way too big to actually put into the context window for a language model. And this brand new architecture of RAGs solved the problem for us. RAGs is just a pattern and you have to actually apply it by using embeddings and vector databases. So for us, we found the magic solution with DataStax. So that we could use a highly reliable vendor to deliver on it. We went from POC to production in a matter of weeks. It was a real simple venture for us.
We found that DataStax was preintegrated into some of the best-in-breed applications already out there such as Llamaindex, and that expedited our journey as we look forward in the coming years of where all this is going. What we see as a marketplace. There are organizations who need agents and assistants and all kinds of AI capabilities. And there's also a lot of service providers who built these and have them to offer what's unclear is how they're going to come together in a safe you know, responsible manner, and we want to be a key piece of that.
Everyone wants to chat to GPT experience but the mega vendors often try and give it to you in one form factor. Imprompt provides a chat to GPT experience to large enterprises so that they can customize it and bring their own solutions to the challenge that we're trying to solve for customers is really empower them to be able to use the new generative AI on their own terms. We want them to be able to use all their existing IT infrastructure and to bring it into the GPT world and not have to worry about the needs of the mega-vendors.
When we started the journey, we looked at Pinecone and Weaviate and some of the other startups in the space. Unfortunately, we did sign with one of them. But we found that they dropped our database unintentionally for unable to recover for six days. With that we decided that it was more important to find a vendor who was more reliable and knew how to move fast. DataStax was the partner because they thought like a professional journeyman, but they were also able to move like a fast startup. We're always concerned about bringing product to market rapidly. And we did so with the DataStax product in very short order.
When we were architecting the product we contemplated how we are going to deal with files generally, a PDF is way too big to actually put into the context window for a language model. And this brand new architecture of RAGs solved the problem for us. RAGs is just a pattern and you have to actually apply it by using embeddings and vector databases. So for us, we found the magic solution with DataStax. So that we could use a highly reliable vendor to deliver on it. We went from POC to production in a matter of weeks. It was a real simple venture for us.
We found that DataStax was preintegrated into some of the best-in-breed applications already out there such as Llamaindex, and that expedited our journey as we look forward in the coming years of where all this is going. What we see as a marketplace. There are organizations who need agents and assistants and all kinds of AI capabilities. And there's also a lot of service providers who built these and have them to offer what's unclear is how they're going to come together in a safe you know, responsible manner, and we want to be a key piece of that.