Everything will be 200 OK :)

Problem We have static resources such as company policies, employee guidelines, portfolios etc that we would like to expose to LLM and offer customers ability for them to query on that information. Since LLM's are pre-trained models they lack information about the content in our own resources. We want to empower our end users with ability to query on our resources for wide variety of use cases. Solution Retrieval Augmented Generation   (RAG) is the solution we have been waiting for to enrich our  application. It can summarized as a two step process.  In the first step, we store our information content through a custom ETL process into a special kind of persistence store named Vector Store where each chunk is stored as a series of multi dimension N dimension vectors through process called Vector Embeddings . In the second step, a semantical similar search is performed between these vectors and the user chat prompt so that it maybe augmented before presented to the LLM for...

Problem In GenAI applications, the LLM's are pre-trained models for information relative to a point in time. This does solve for cases when certain prompts are made for that information. However, they fail to provide adequate response to request for information outside its knowledge bank. The demand for this can vary per business use case and the problem we trying to solve to provide that information to LLM. Information can be static such as a resource, or a dynamic information obtained from your own knowledge bank from business information to wide variety of application use cases.  Solution For static resources such as document, we have a solution that we will discover in upcoming posts. For dynamic information such as retrieving information from our database or a custom business logic where we return the data back to the LLM for it to make sense of it, we can use a concept termed Tool Calling.    In its simplest form this is merely your custom business logic done as a f...